July 15, 1914 
Influence of Host on Puccinia 
305 
each have been noted. It is, therefore, with much confidence that the 
writer sets forth the facts as to the limited spread of each of these rusts 
in the uredinial stage. 
The same thing is true to even a greater extent with Uromyces andro- 
pogonis. Only one time has the writer been able even to carry this rust 
from its aecial stage back to the Andropogon, and then only one or 
two sori were produced. 
A careful study of the fresh urediniospores of both Puccinia ellisiana 
and Puccinia andropogonis under the microscope seems to indicate that 
the epispore in both species is either slightly viscid or slightly gelatinous; 
and, if this be true, it would explain the inability of the urediniospores 
to spread the rust to distant stools of Andropogon. However, the writer 
is not at all certain that the epispore when fresh has this viscid character, 
but the fact is fully established that in this region neither rust will spread 
very far from its aecial host. In fact, the writer has yet to see a stool 
of infected grass which had certainly been originally infected by uredinio¬ 
spores. In every instance the infected stools found even in the fall of 
the year were sufficiently close to the secial host for the aeciospores to 
have been the only and sole infecting agents. In this respect these 
rusts differ markedly from the common grain rusts, Puccinia rubigo~vera 
and Puccinia graminis , which are able to spread over large areas from 
the urediniospores alone. 
CHARACTER AND SOURCE OE CULTURE MATERIAL 
In the culture work of 1913 the teliosporic material used was not 
pedigreed material from inoculations made under control conditions in 
the greenhouse, but was material grown for three years under the writer's 
direct supervision on a plat of ground near his residence in Clarendon, 
Va. The aecial host each year was Viola sagittaia t and the telial host was 
Andropogon mrginicus. That there are no Pentstemon plants within a 
radius of a mile of this place was determined by a careful search each 
year for the last live seasons. The plat of ground used was a glade sur¬ 
rounded on all sides by woods consisting of oaks (Quercus spp.) and 
pines ( Pinus spp.). 
The culture work of 1913 seemed to indicate that the species of the 
aecial host might influence to some extent the ability of Puccinia ellisiana 
to infect Pentstemon. For instance, teliosporic material whose aecial 
host was known to be Viola sagittata infected Pentstemon, while teliosporic 
material with V. papilionacea as its aecial ancestral host did not. In 
1914, however, telial material from either aecial host readily infected the 
Pentstemon plants. The failure in 1913 of the teliosporic material which 
had V. papilionacea as its aecial ancestral host to infect Pentstemon was 
probably due to two things: 
(1) The first series of inoculations with the rusts from this aecial host 
was made too early. The Pentstemon plants used had not yet reached 
their susceptible period. 
