390 BROOKLYN BOTANIC GARDEN MEMOIRS 
other cultures of the organism were used to inoculate the varieties a 
different arrangement of susceptibility became evident. Varieties 
quite immune to the one strain were severely attacked by another 
strain. All the varieties proved quite susceptible to at least one of 
the strains used. This indicates the existence of distinct races of this 
parasite with fairly definite host limitations. 
Edgerton and Moreland (33) have made similar studies with cul- 
tures of this same fungus. Their results also indicate. differences in 
the infecting capacity of strains of the fungus isolated from different 
varieties of beans. Some beans, like the snap beans, appear to be 
quite susceptible to a number of different strains. Other varieties, 
while quite susceptible to certain strains, are resistant to strains from a 
different source. 
Edgerton and Moreland have also studied cultures of Glomerella 
gossypii (South.) Edg., the cotton anthracnose fungus. They do not 
find evidence for the existence of specialized strains in this fungus, 
for all the cultures isolated from different sources were able to infect a 
large number of cotton varieties. 
GENERAL DISCUSSION—BRIDGING Hosts 
The above review of the investigations with reference to host 
specialization of parasitic fungi indicates that the phenomenon is of 
general occurrence. The work done is particularly extensive in con- 
nection with the rusts and the powdery mildews, but sufficient has 
been accomplished in other groups to make clear the presence of 
specialized races. 
It is highly probable that the same phenomenon is of wide occur- 
rence among other groups of plant parasites. A large number of 
so-called species of the Imperfect Fungi, as Cercospora, Phyllosticta, 
and Septoria, may really be only specialized races of a relatively small 
number of forms distinct on structural grounds. In several genera of 
the Imperfect Fungi, as those mentioned, a very large number of 
species have been recorded; in fact many of them are, mainly at present 
at least, identified by the host upon which they grow. Cultural ex- 
periments may result in grouping many of these together, at the same 
time making clear the physiological host relations. 
We are not, however, to assume that host specialization is of 
universal occurrence. There are several cases on record where the 
fungus shows no evidence of the specialization of its hosts. A striking 
case of this sort is that of Puccinia subnitens Diet. Arthur (5) has 
been able to infect ten hosts, belonging to the families Chenopodiaceae, 
Cruciferae, and Capparidaceae, with teleutospores from Divstichlis 
spicata. Bethel (18), using teleutospores from the same grass, recently 
