3°4 
Journal of Agricultural Research 
Vol. II, No. 4 
One patch of grass infected by Puccinia andropogonis has been ob¬ 
served by the writer for four years, and during this time the size of 
the infected area has not increased, though there are many stools of 
Andropogon near the infected area. In the center of this area are 
several infected groups of plants of Pentstemon laevigatus, while at a 
distance of 40 feet from these there are other plants of the same species. 
Yet during the last four years the rust has not crossed this 40-foot gap. 
Eight feet from the infected Pentstemon plants no signs of any rust on 
the grass can be found, even in the fall of the year after the rust has 
had all summer to spread. If the rust had ever crossed over the 40-foot 
gap to the grass near the uninfected Pentstemons, it certainly would have 
perpetuated itself, since both aecial and telial hosts were present and 
adjacent to each other. The rust had ample opportunity, as far as 
contiguity of telial host and proximity of infected plants is concerned, 
to be carried this distance. Nevertheless, it was limited to the stools 
of grass immediately adjacent to the infected secial host, the Pentstemon. 
Likewise the seciospores and urediniospores of Puccinia ellisiana are 
not carried over any great distance by the wind and can only infect 
stools of Andropogon which stand within about 6 feet of the secial host. 
This has been previously noted by the writer. 1 
The difficulty with which the urediniospores of either Puccinia andro¬ 
pogonis or Puccinia ellisiana infect other stools of Andropogon was 
clearly shown when the writer attempted to obtain a large amount of 
teliosporic culture material by setting uninfected pots of Andropogon in 
actual contact with the pots containing the grass already infected from 
the seciospores. The experiment was a failure, as no infection occurred. 
The uninfected stools of grass remained free of the rust, even when the 
tips of the blades intermingled with the blades of the infected stool. 
Check stools situated 10 to 20 feet from the infected stools and planted 
m the same cold frame showed no infection. 
Tor two years the writer has attempted to transfer the rust from 
stool to stool by inoculating with the urediniospores under bell jars in 
the greenhouse, but has been unsuccessful, even under such favorable 
conditions. This failure to infect with the urediniospores in the green¬ 
house may have been due to the high temperature and the condition of 
the blades of grass. It is probable that in nature these rusts are able 
to infect adjacent stools of grass on whose leaves the urediniospores 
fall, but to date the writer has been able to obtain infections on Andro¬ 
pogon only by means of the aeciospores. These facts are given in detail 
to show the improbability of either of these two rusts being mixed in any 
of the culture material used for the experiments performed in 1913 and 
1914. 
For five years the writer has been studying in field and laboratory 
this group of Andropogon rusts, and during that time the peculiarities of 
1 IfOng, W. H. Op. cit., p. 170. 
