220 



AMERICAN JOURNAL OF BOTANY 



[Vol. 9 



TABLE 4 



The results cannot be considered as entirely convincing in view of the 

 many factors involved in a successful inoculation. The evidence indicates, 

 however, that, as to the minimal number of uredospores which can possibly 

 produce infection, it is possible for a single uredospore to produce infection. 

 In each of the two cases in which infection was produced after inoculation 

 with a single uredospore, the rust appeared after the usual incubation period 

 as a very small pustule at the point of inoculation. In one case the in- 

 fected leaf withered shortly after the appearance of the pustule; in the other 

 instance the mycelium showed normal growth, and about a week later 

 seven new uredosori were produced in a ring around the first pustule. 



As regards the second question, the minimal dose of uredospores which 

 will certainly produce infection, the data indicate this to be, for the con- 

 ditions under which the work was done, between 100 and 125. This is 

 high. In view of the varying viability of spores taken at one time from the 

 same pustule, and of their further variation with the age of the pustule, 

 it was not considered possible actually to test the germination for each 

 sample used. However, the spores were always taken from the surface of 

 a large and pulverulent sorus, and, considering that the index of germination 



