Specialization of Parasitism in the Erysiphaceae. 113 
nut ion of peritbecia among powdery Oidium- patches took place 
on the twenty-fifth day (October 5) after the inoculation of the 
leaves. 
The Oidiwn on Plantago lanceolata proved unable to infect P. 
major and Taraxacum officinale (see Table 7). 
In certain experiments with the Oidium of E. Cichoracearum 
on Plantago major (Table 8) a production of perithecia resulted on 
inoculated leaves of P. major and P. media. In the first experiment 
(No. 199) seedling plants, six weeks old, of P. major and P. media 
were used; three leaves of the former species, and two of the 
latter, were inoculated. On the tenth day after inoculation, 
numerous patches of mycelium bearing a large number of clustered 
conidiophores were visible on all three leaves of P. major ; while one 
leaf of P. media bore a few scattered conidiophores. On the fifteenth 
day two of the leaves of P. major bore young (yellow) perithecia 
among the powdery Oidium- patches; the one infected leaf of 
P. media was now clearly fully infected, and bore a large number 
of conidiophores. On the nineteenth day numerous perithecia—some 
turning brown—were observable among the still densely powdery 
O/rfno^-patches on all the three leaves of P. major ; on the 
infected leaf of P. media one young (yellow) perithecium was now 
visible. By the twenty-fourth day the fungus had formed hundreds of 
brown perithecia on the infected leaves of P. major ; on P. media 
the single perithecium that was produced had now turned dark- 
brown, and on examination proved to contain young asci. It may 
be noted that the single leaf of P. media which became infected 
was a cotyledon, and that although evidently full infection occurred, 
the number of conidiophores produced was not so great as on P. 
major . In the second experiment (No. 229) two leaves each of 
seedling plants of P. major (seven weeks old) and P. media (ten 
weeks old) were inoculated. On the sixteenth day after inoculation 
(when the experiment was discontinued) the two inoculated leaves 
of P. major bore large powdery ChW»<//i-patches ; while one of the 
leaves of P. media was fully infected, and bore numerous nearly 
powdery patches of conidiophores. In this case the infected leaf 
was not a cotyledon, but the second or third leaf. In other 
experiments, in which eight leaves of P. media were inoculated, no 
infection resulted. No infection followed the inoculation of P. 
lanceolata, Galium Aparinc, and Eupatorium cannabinum (see 
Table 8). 
The results of the above experiments show that the form of 
