493 
Rust of Bromes ( Puccinia dispersa). 
It has been previously pointed out by Marshall Ward that 
the infection capacities of the brown rust spores are profoundly 
affected by conditions of formation and also by the host upon 
which they fall, and furthermore, that the optimum condition 
arises from infecting a species with spores formed on that 
species, which infection is least likely to succeed with spores 
of a species on those farthest removed in systematic position. 
The results of series A are negative and valuable only in 
connexion with further work on these species. In series B 
one would expect to find species with at least not very remote 
relationships with B. mollis. 
Such species as B. commutatus , B. inermis ( arvensis ?), and 
macrostachys may have closer affinities with one of the other 
groups of Brome species than with the mollis group, but no 
close affinities with the B. sterilis. It will be seen by an 
examination of the members of the series that the expecta- 
tions as to systematic position are in general realized. That 
intermediate forms between B. mollis and B. sterilis exist 
is an inference which may perhaps be drawn from series 
D; i.e. at least B. Gussoni , B. Krausei and B.pendulinus are 
quite susceptible to infection with spores from both species. 
B. Gussoni leans notably toward the sterilis side, while B. 
pendidinus and perhaps B. Krausei and B. vestitus would 
seem to be nearer to B. mollis. B. Krausei is remarkable in 
the case of infection from both species, and there is probably 
some other factor to be considered besides the morphological 
and biological factors involved in the systematic position, 
because if B. Krausei is an intermediate form, one would not 
find such extraordinary susceptibility to either species as is 
actually shown to exist toward mollis spores. If this ad- 
ditional factor is not all-important, and there is no reason to 
believe such to be the case, one would seek in the mollis group 
for the closer affinities of B. Krausei. 
The above results are all those of averages and there are 
several possible sources of error which must be borne in mind. 
The .chief of these is to be found in the possibility of the 
admixture of stray spores from other species of Bromus than 
