152 
Journal of Mycology 
[Vol. 12 
influence of the substratum in the case of parasitic plants, how¬ 
ever, is doubtless largely comparable with the influence of the 
substratum in non-parasitic plants, only more pronounced. 
But there is an influence which has helped to delimit both 
species and genera among the Uredinales, not found apparently 
in any other group of plants. This influence is difficult to 
define, but it is manifested in directing the phylogenic course 
of development within the group, by which the life-cycle is 
shortened. 
There appears to be ample justification in assuming that 
at a comparatively early period in the evolution of the Uredinales 
all the rusts possessed four forms of spore-structures, pycnia, 
aecia, uredinia and telia. We need not go back of this period 
to inquire how they came to have the four kinds of spores, as 
it does not affect in the least our present contention. But dur¬ 
ing the universally four-sporecl condition for the order, some 
influence began to affect the relative production of the several 
spore-forms, which eventuated in the suppression of one or more 
of these from certain species or group of species. As a result 
of this influence we find that the Uredinales of the present day 
consist of groups of species in the several divisions of the order, 
generally regarded as genera, which in many cases can be again 
separated into at least four groups of species, according as they 
possess all spore-forms, all but aecia, all but uredinia, or finally 
only telia. In each of these four groups the pycnia are gener¬ 
ally present, but in some species of the last named group even 
pycnia may fail. 
Among the melampsoraceous rusts there has been an un¬ 
premeditated, and largely unrecognized tendency to regard the 
absence of certain spore-forms as among valid generic charac¬ 
ters, but among the pucciniaceous rusts such a tendency is dis¬ 
tinctly opposed. That this is, however, a generic character of 
importance among rusts generally, I not only venture to affirm, 
but believe that it will in many cases be found to be associated 
with other characters further indicating true generic grouping. 
It is among the genuine Pucciniae after other genera have been 
removed as indicated above, that this character for separating 
genera finds its most uniform and conspicuous application. 
This is in fact exactly what should be expected, for this group 
undoubtedly represents the highest and most complex develop¬ 
ment of the Uredinales. 
In advocating the importance of recognizing the life-cycle 
in drawing generic distinctions it is well to forestall misappre¬ 
hension by pointing out that the usual absence of a spore-form 
does not necessarily constitute an abbreviated life-cycle. Many 
species of rusts in northern regions especially have the habit 
of propagating themselves from year to year largely by the ure- 
diniospores which survive the winter, either as continuously pro- 
