24 



INTRODUCTION 



[CH. 



normally immune from the attacks of the biological form used. Thus if 

 conidia oi Erysiphe Graminis growing on wheat, are sown on uninjured leaves 

 of wheat and barley, the result is the infection of the wheat but never of the 

 barley; yet the conidia grown on wheat readily infect the uninjured surface 

 of a cut or burnt barley leaf; in the case of the burnt leaf, where the whole 

 thickness of the leaf in the burnt region becomes discoloured and apparently 

 dead, the mycelium is found on the living cells which border the altered 

 patch. Since the cuticle, hairs and other anatomical characters of the epi- 

 dermis on which the spores are sown are not affected by the treatment of 

 the cut leaf, it is clear that resistance does not depend on such factors ; it 

 must be referred, as in the case of rusts, to the physiological condition of the 

 cells or of their contents. In nature injuries caused by insects are sufficient to 

 destroy in the same way the resistance of the potential host. 



These facts have a practical bearing since diseases on the weed grasses 

 surrounding a field of corn, even if they are not able directly to affect the 

 uninjured leaves of the crop, may establish themselves on injured tissues. 



The various hosts of a given morphological species of fungus differ very 

 much in their susceptibility to infection by the different biological forms of 

 the parasite; thus Bromus racemosits, though markedly susceptible to its 

 own form of Erysiphe Graminis^ is completely immune against infection by 

 conidia of the same species grown on B. commutatits, B. interrnptus, B. velu- 

 tinus and others. This immunity is particularly remarkable in the case of 



conidium on 

 commutatus 



conidium on 

 secalinus 



J^ racemosus j^_^ 



conidium on 

 velutinus 



conidium on 

 interrupius 



conidium on 

 hordeaceus 



the conidia from B. commutatus, since B. racemosus is morphologically so close 

 to B. commutatus that it is regarded by most systematists as no more than 

 a variety of that species. 



conidia on 

 arduennensis 



conidia on 

 commutatus 



conidia on 

 secalinus 



conidia on 

 adoensis 



