ARTHUR: PROBLEMS IN THE Stupy OF PLant Rusts 9 
outside of a single tribe. No full explanation of this is yet forth- 
coming, but when found it is likely to throw a flood of light upon 
the true nature of parasitism. 
We return from this digression to note that Eriksson has 
vigorously attacked the difficult problem how the rust is carried 
through the winter and again becomes a menace to the farmer in 
northern regions, where the mycelium is not perennial and where 
barberry and alkanet or other suitable host plants are wanting on 
which the cycle of development can be completed. In trying to 
account for the facts he has elaborated and widely published a 
micoplasma theory: the symbiotic and intimate association of the 
protoplasm of the host and its parasite, unseparated by cell walls. 
This theory has not been accepted by most botanists, and since 
the recent publication of Marshall Ward's researches upon the 
rusts of the genus Bromus, appears to have no authentic or even 
probable facts upon which to rest. The studies of Carleton upon 
the American Euphorbia rust, in which it was found that the 
mycelial fungus may be carried over the winter in the seeds, may 
throw some light upon Eriksson’s anomalous results, without 
invoking a micoplasma theory. We may, indeed, believe that to 
solve the practical problems of the mode of propagation and the 
power of infection of the several grain rusts requires all the astute- 
ness and skillful experimentation that can be brought to bear ; 
and the repeated failures of the ablest investigators only empha- 
size the difficulties. 
The expansion of knowledge through DeBary cultural methods 
cannot be better illustrated than by citing the results attained by 
Dr. Klebahn of Hamburg with the willow and poplar rusts. This 
investigator stands foremost among those who have used each sea- 
son’s work asa basis for shaping the work of the following season, 
and who pari passu with cultural results have studied morpholog- 
ical characteristics, combining the two classes of data into taxo- 
nomic expression. Most systematists have contented themselves 
with listing the willow rusts under J/elampsora saltcina and the 
poplat rusts under Jle/ampsora populina, or equivalent names. 
This was the disposition of them made by Winter and Burrill. 
Some have essayed segregation, but with an unsteady hand. 
‘ Thimen in 1879 described seven species of willow rusts and 
