122 
FRANK DUNN KERN 
sider only the relation between a parasite and its host, or hosts when 
it passes to several allied forms, as indicated in the foregoing paragraph. 
No matter how difficult this may appear to be we are obliged to desig- 
nate this as simple parasitism. To contrast with this there are the 
cases of parasites which are known to inhabit wholly different and 
unlike hosts in different stages of their life histories. Such a surprising 
and, as it seems, unnecessarily complex parasitism exists in^both ani- 
mal and vegetable kingdoms. In those species which change hosts, 
and on that account are termed heteroecious, there is frequently more 
than one method of propagation. Protozoan forms which come under 
this category are said to multiply by schizogony, a kind of multiple 
fission, and by sporogony, a process of spore-formation preceded by 
conjugation. Sporogony usually either makes possible the invasion 
of a new host or takes place only in connection with the second or in- 
termediate host. The evident food relations between the hosts of 
many of the heteroecious animal parasites, such as the mosquito and 
man and various mammals in the case of malarial forms, renders the 
association of the unlike hosts not so difficult to appreciate. 
In the higher fungi, especially in the group popularly known as the 
rusts (Uredinales) , there are many instances of heteroecious parasi- 
tism and in none of them is there the slightest hint of a relationship 
between the two hosts. There are many forms inhabiting coniferous 
trees and from these they may pass over in their alternate stage to 
plants widely separated in taxonomic classification. Thus we may 
find a form alternating between a fir and. a fern, a cedar and an apple, 
or a pine and a composite. It seems impossible to escape the con- 
viction that some extraordinary relationship must exist between these 
two sets of hosts occupied by a heteroecious form and yet in the half- 
century since we have had definite proof of heteroecism in the rusts no 
satisfactory explanation has been forthcoming. In North America 
we have today almost exactly one hundred demonstrated cases of 
heteroecious life-cycles in this group. The rusts are of especial 
interest from our present point of view not only because of heteroecism 
but also because of the existence of autoecism within the same group, 
and the consequent opportunity to compare these two types of adap- 
tation. Heteroecious forms always possess what may be termed a 
long life-cycle divided necessarily into two phases, on the one host 
consisting of telia, either alone or accompanied by uredinia, and on 
the other host of aecia, preceded by pycnia. Telia, uredinia, and aecia 
