150 
Journal of Mycology 
[Vol. 12 
groups of species. One of the impediments at the present time 
to an understanding of the interrelationship of rusts lies in the 
lack of reasonable segregation of genera. In support of this 
statement one need only recall the fact that the genus Puccinia 
as now consituted contains more than half of all known species 
of rusts, and what may not be so well known, that within this 
category are contained groups of the most diverse forms and af¬ 
finities. To be assigned to this genus requires only that the rust 
shall possess a two-celled, stalked teliospore. No consideration 
need be given to the nature of the spore wall, whether homogene¬ 
ous or differentiated into well marked layers, or to the 
number and position of the germ pores in each cell, or to the 
question of simple or compound stalks. It is also unnecessary 
to ask whether the life cycle of the rust possesses pycnia, aecia 
and uredinia, in addition to the telia, or whether one or more of 
these is wanting, or what may be the origin of sori in any or all 
of these four stages in relation to the substratum. Yet all these 
characters, and some others, should be kept in mind to rightly 
appreciate the validity of a genus in the Uredinales. In short 
the genus Puccinia is founded upon what is essentially a single 
character, which can not be shown to be correlated with other 
characters going to form a natural grouping of closely related 
organisms. A very similar genus is that of Uromyces , which 
differs from Puccinia apparently only in possessing one-celled 
instead of two-celled teliospores, and all that has been said of 
Puccinia aplies with equal force to Uromyces. These two genera 
.are fine examples of the heterogeneous results of founding genera 
on a single technical character when it can not be shown to be 
also a representative character. 
Let us now turn from the negative to the positive side of the 
discussion, and instead of insisting upon the artificial construc¬ 
tion of the genera Puccinia and Uromyces, let us see what segre¬ 
gation can be made of the species to show more fully their affini¬ 
ties. First of all it will be necessary to study more fully than is 
usual both spores and sori of all the stages or phases of each 
species, including the pycnia. Our attention will soon be at¬ 
tracted to the fact that while the pycnia of the majority of 
species are flask shaped, and arise under the epidermis, some are 
conical or flat, and arise under the cuticle. We shall further find 
that these differences are correlated with characters in the other 
spore forms, especially in the spores and sori of the uredinia 
and the spores of the telia. 
Removing now all species with subcuticular pycnia, and 
directing attention more particularly to the uredinia of this 
segregated group, we shall find species in which the uredinio- 
spores are distinctly different at apex and base, reminding one 
of the urediniospores of the genus Ravenelia. Correlated char¬ 
acters will be found to set aside a group of genera having be- 
