Journal of Agricultural Research 
Vol. II, No. 4 
318 
markings on the walls of the spores) have been changed by the new 
aerial host, Pentstemon. The spores have changed from thick to thin 
walls, from verruculose to spinulose, from ellipsoid to globose, from 16 
to 19 by 21 to 23 jj, to 20 to 24 by 20.8 to 25.6/1, from walls with 15 to 25 
warts across the spore to walls with 12 to 14 spinules across. In every 
instance the Viola rust has changed its characters to those of the ordi¬ 
nary Pentstemon rust. In the teliospores the same trend away from 
the characters of the typical Viola rust and toward those of the ordinary 
Pentstemon rust is seen. The characters of the teliospores of “Puccinia 
ellisiana from Pentstemon” are more nearly intermediate between the 
two regular rusts than are the characters of any of the other stages, but 
the differences in the characters of the teliospores of the typical Puccinia 
ellisiana and Puccinia andropogonis are so slight that one is usually not 
certain which rust he has unless the urediniospores are present. In 
other words the characters of each successive stage of Puccinia ellisiana , 
when it has Pentstemon for its aerial host, change to correspond to those 
of the ordinary Pentstemon rust, Puccinia andropogonis. 
The large number of successful cultures made on Pentstemon with 
Puccinia ellisiana , the vigor and abundance of the infections obtained, 
the character of the culture material used, the many sources from which 
the culture material came, the use of pedigreed culture material grown 
under control conditions in the greenhouses, the special care taken in 
the actual culture work to avoid accidental contamination, and the 
duplication this year of last year's culture results, all prove conclusively 
that the results obtained in these experiments were not due to accidentally 
contaminated culture material—that is, to telial material containing 
viable spores of both rusts—but were due to changes produced by the 
aerial host through which the rust passed. 
The infection of Viola spp. by Puccinia andropogonis and by “Puccinia 
ellisiana from Pentstemon” further corroborates the results obtained 
with Puccinia ellisiana. These experiments undoubtedly show that the 
ordinary Pentstemon rust, Puccinia andropogonis , can be produced 
from the Viola rust, Puccinia ellisiana , by simply passing Viola rust 
through Pentstemon as an aerial host. This process is so easy and the 
infection so vigorous and abundant that it certainly can and does occur 
in nature, thus originating the ordinary Pentstemon rust. But the 
reverse process, the passing of the regular Pentstemon rust through the 
Viola spp. and thus back to Puccinia ellisiana , is so difficult to accomplish 
even under the most favorable conditions that it seems probable that 
such a process would rarely, if ever, occur in nature. 
Puccinia ellisiana and Puccinia andropogonis are then but different 
forms of the same species, since both can be produced from the same 
telial ancestor. 
