Erysiphe Polygoni. G. 0. Searle. 279 



This table indicates that the form of Erysiphe Polygoni on 

 Trifoliimi pratense is speciahsed on that species and is unable 

 to infect other species of the same genus or other genera of 

 the same family, thus, as far as the experiment was carried, 

 confirming the more extensive series carried out with the oidium 

 on T. pratense by Salmon (4). 



Two further experiments with clovers were tried, in one of 

 which conidia from Trifolium minus were found unable to 

 infect T. pratense or Onohrychis sativa, and in the other conidia 

 from Trifolinm hyhridum were found unable to infect T. repens. 



A short series was then tried with the form on Pisum sativum. 



Table IV. 



Inoculations with conidia from Pisum sativum. 



No. of No. of 



Hosts inoculated inoculations infections 



Pisum sativum ... ... ... 5 — 



15 +10 



Onobrychis sativa ... ... 10 o 



Vicia Faba ... ... ... 24 o 



Trifolium pratense ... ... 4 o 



This table confirms the fact, already noted by Salmon, that 

 the oidium on Pisum sativum should rank as a "biologic form." 



A single experiment with the oidium on Capsella Bursa- 

 pastoris showed that the conidia, though able to infect fully 

 Capsella Bursa-pastoris (six inoculations), were unable to infect 

 the Swede (six inoculations). 



In confirmation of Table I a further experiment with the 

 conidia from Rape was tried ; in this case eight inoculations on 

 Turnip caused six full infections, but ten inoculations on 

 Kohl-rabi produced no infections, whereas one would have 

 expected that " subinfections " would have resulted. 



All the above inoculations were carried out between July 29th 

 and Sept. 26th after which date the spread of natural infection 

 in the greenhouse was so general that further trials were 

 entirely vitiated. 



A number of series was then commenced in the laboratory 

 employing excised leaves placed on moist filter paper in Petri 

 dishes, and pieces of stem placed in larger dishes. These experi- 

 ments gave some interesting results especially in their bearing 

 on the growth of species of the Erysiphaceae on internal tissues 

 of plants in contradistinction to their usual ectoparasitic 

 existence. This phenomenon had already been noted by 

 Salmon (10) when working with Erysiphe graminis. 



Before describing these series it will be necessary to deal 



