y" 



Erysiphe 207 



species is confined to certain host plants belonging to the Labiatae. 

 In the size, etc., of the perithecia, nature of appendages, number 

 and size of asci, etc., there is no difference— as indeed is generally 

 admitted — between E. galeopsidis and E. cichoraccanim, so that the 

 claim of the former to rank as a distinct species rests on two sup- 

 posed separative characters, viz.: the lobed haustoria and the non- 

 development of spores on the living host-plant. 



This latter peculiarity is found also as a rule in E. gnuninis, 

 but we must notice that in this species spores are sometimes pro- 

 duced in late summer or autumn, 'so that in E. gmminis, at least, 

 it is seen that this character is not one of specific importance. 



So long, however, as the absence of spores and the presence 

 of lobed haustoria could be considered as correlated characters, 

 and ones not occurring in E. cichoraccarum, E. galeopsidis could 

 be maintained as a distinct species. Some cases I have examined, 

 however, make me doubt if we can consider either of these char- 

 acters as absolutely characteristic of E. galeopsidis. In the exam- 

 ination of a large amount of material I have certainly found that 

 the Erysiphe on species of Eamium, Galeopsis, Stachys, etc., in- 

 variably contain asci without any trace of spores, and in all cases 

 where examination was made the mycelium was found to possess 

 lobed haustoria. But, on the other hand, as regards this character 

 of the non-production of spores, there are forms of E. cichoracc- 

 anim which appear to be connecting links. Such are the forms 

 on Mentha arvensis, M. aquatica, Lycopiis Europaeus and Scutellaria 

 lateriflora. On these host-plants, some specimens of the fungus, 

 apparently mature, have perithecia in which the asci show no trace 

 of spores, and are in fact indistinguishable, as regards perithecia, 

 from E. galeopsidis ; in other specimens two more or less well- 

 formed spores are found in the ascus. The haustoria of the fun- 

 gus on these hosts are not lobed (although there is sometimes a 

 tendency for them to have a crenulate margin), and for this 

 reason, probably, De Bary placed the fungus on Mentha and Ly- 

 copus under E. cichoraceannn. The fungus on Scutellaria ver>' 

 rarely shows any signs of the formation of spores in the ascus, 

 and it is probably on this account that Burrill places it under E. 

 galeopsidis ; the haustoria, however, are not definitely lobed. An- 

 other striking case Is that afforded by the fungus on species of 



