6o 
Parasitism in the Erysiphaceae 
fungus occurring on B. racemosus , and inoculated with them twenty- 
two leaves of B. “ hordeaceus." In seven days twenty of the 
twenty-two leaves bore patches of mycelium with conidiophores and 
little powdery masses of spores. These spores, produced on B. 
“ hordeaceus" by inoculation with spores from B. racemosus, were 
sown on a leaf of B. commutatus, and caused full infection in eight 
days. This experiment, which w’as safeguarded by controls, gives 
definite proof that certain species of Bromus serve as “ bridging 
species.” 1 (These results are illustrated in fig. 65). 
There is little doubt that when all the details of the inter¬ 
relations between the forms of the fungus and the various species of 
Bromus are worked out, the species of host which serve as meeting- 
places or as bridges will be found to be very numerous, and that 
consequently many of the “ biologic forms ” of the fungus within 
the genus Bromus which now appear to be distinct, will be found to 
merge into one another at these points. 
1 The same kind of inter-relations between a parasite and cer¬ 
tain hosts has been discovered by Freeman (Annals of Bot., 
XVI, 487-494, 1902), and Marshall Ward (Annal. mycolog. 
I., 132-151, 1903), in the case of one of the XJredineae (Puccinia 
dispersa, Krikss.) on species of Bromus. 
R. MADt.EY, PRINTER, WHITFIELD STREET, LONDON, W. 
