Relation between Host and Parasite 19 



sufficiently marked to prevent the forms from being considered as 

 belonging to the species. Or (2) We must suppose that the coni- 

 dia of 5. hnmuH on a strange host-plant, although producing a 

 mycelium capable of giving rise to conidia, might be unable to con- 

 tinue its development and produce pcrithecia, in which case the 

 form would quickly die out. Certain Oidium-{ox\\\s, e. g., that on 

 species of Myosotis, appear seldom, if ever, to produce perithecial 

 fruit. Finally, (3) That the conidia might have ultimately produced 

 perithecia similar to those of ordinary S. hiimitli. This seems 

 perhaps unlikely, when we remember that S. huinnli has never 

 been recorded on Taraxacum, notwithstanding that this plant is 

 a common weed in many places {c. g., hop-gardens) where 5. 



IluuiuH grows. 



The compilation of a list of the host-plants of the Erysiphaccac 



is attended with many difificulties. 



I. In the first place the fungus has frequently been wrongly 

 determined in Exsiccati. An instance will make this clear, and il- 

 lustrate how wrong plants may become recorded as hosts for certam 



pecies 



Sph 



Castagnei" is given as the species occurring on these hosts : 

 Inula hirta (Sacc. Myc Ven. 630). Plantago maritima (Rab. 

 Fung. Eur. 19 16), P. w^y'^r (Rab. Fung. Eur. 104S), Achillea Ptar- 



) 



Eupatoj 



ed. 2, 591, 592), Trifolium medium (Syd. Myc. March. 3052), 

 Calystegia ' scplum (Syd. Myc. March. 432), Catalpa syringaefolia 

 (Syd. Myc. March. 1640), Falcaria vidgaris (Syd. ^^lyc. March. 

 1 541) and Stachys alpina (Sacc. Myc, Ven. 149 1). In the first 

 six specimens the flmgus is Erysiphe cichoraccarum, in the next 

 four E. Polygoni and in the' last E. galcopsidis. Here we have 

 eleven species of host-plants wrongly given for one species in 

 specimens sent out in exsiccati. It is, moreover, probable that 

 the error does not stop here, but that it is perpetuated by the 

 fundus on these hosts being named and recorded as " S. Castag- 



VUO KJi^ Lliv^o^. .IV^.-..^ "V-";to 



nci'' without examination by subsequent collectors ; it is significant 

 that four of the plants given above are stated to be hosts of. 5. 



Castagnei by several authors. 



2. The list of host-plants has been further complicated by the 



