> 
REED: SPECIALIZATION OF PARASITIC FUNGI 349 
Whatever term has been applied, the underlying conception has 
been the same, namely, that these races, strains, forms, etc., of distinct 
morphological species of fungous parasites differ, not in discernible 
structural features, but in their physiological behavior, as indicated 
by their ability to infect some hosts and not others. They differ in 
their ability to establish the parasitic relation with particular hosts and 
thus secure the necessary food for their normal development. The 
phenomenon is distinctly physiological and is doubtless quite com- 
parable to the well-known behavior of saprophytic fungi on different 
chemical substrata. Various saprophytes, structually similar, vary 
in their ability to utilize different chemicals as sources of food, de- 
pendent on their capacity to secrete the necessary enzymes. While 
the strains of parasites may differ essentially in their ability to secure 
food from a particular host, we must keep in mind the possibility of a 
more complicated series of relations in which toxin and antitoxin 
production are involved. 
Many investigators of the phenomenon of host specialization have 
made a large number of species on the basis of the results of their 
inoculation tests. This is especially the case in the rusts where 
Klebahn, Eriksson, Schneider, Fischer and others have raised many 
forms to specific rank, although no distinct structural differences can 
be observed. It may be noted that the races of Puccinia dispersa, 
P. sessilis, P. Ribesit-Caricis, P. extensicola, Coleosporium Campanulae, 
Melampsora populina, M. Tremulae and others, referred to below, are 
regarded as good species by some students. 
Fischer (45), in connection with the rusts, accepts as species the 
following: 
1. All rusts which are structurally distinct. 
2. All rusts which have a different life-cycle; for example, forms 
which are distinguished by the presence or absence of certain 
spore-forms. 
3. All forms which differ in their choice of hosts, in so far as the 
hosts belong to different genera. In heteroecious rusts species 
are recognized when the hosts of one generation, aecidial or 
uredo and teleuto, belong to two different genera. 
Fischer unites under one species as formae speciales or specialized 
races all rusts which differ only physiologically and whose hosts are 
species of a single genus. Whether a particular rust is a physiological 
species or a specialized race is thus determined by the range of its 
hosts. 
It is doubtless true that many rusts, and other parasitic fungi 
as well, which can be distinguished only by the hosts upon which they 
grow, are just as distinct forms as others which are characterized by 
24 
