REED: SPECIALIZATION OF PARASITIC FUNGI 401 
range of varieties than P. coronata avenae. ‘The former, as noted above, 
occurs on a number of species of grasses belonging to different genera, 
whereas the latter is closely restricted to the genera Avena and Ar- 
rhenatherum. 
It has been suggested that there is a connection between stability 
of host species and the occurrence of specialized races. Edgerton and 
Moreland (33) suppose that the explanation of the difference between 
Bean Anthracnose and Cotton Anthracnose is due to the fact that 
bean varieties are quite distinct and well marked, with little or no 
crossing and, accordingly, no intergrading forms. On the other hand, 
cotton varieties readily cross and thus a series of intergrading forms 
occurs onto which the anthracnose fungus may spread. This explana- 
tion, however, cannot have any very general explanation. The evi- 
dence indicates that the willows hybridize quite readily and yet a 
number of specialized races of willow rusts are recorded. 
Magnus (96) suggested that the existence of these specialized races 
may indicate an adaptation on the part of the parasite to live on 
particular hosts. He makes a distinction between adaptive races and 
biologic forms. The former term is applied to strains or races of a 
parasite which tend to infect certain hosts more readily than others. 
On the other hand, the biologic form is sharply restricted to its hosts. 
Dietel (30) points out that a parasite may first have attacked a wide 
range of hosts, gradually becoming broken up into races adapted to 
certain hosts and finally limited to them. It is possible that the rela- 
tive abundance of the hosts may have been a factor in this process. 
The presence or absence of hosts in a given locality may also have 
played a part. 
The question has often been raised as to whether these specialized 
races of fungi are constant or fixed. Montemartini (100) has recently 
raised the question again as a result of his studies with certain para- 
sites. He inclines to believe that these races are not fixed and definite 
and so capable of being carried from one region to another but rather 
that they are local adaptive forms, perhaps dependent on the vari- 
ability or distribution of the host plants. He suggests that they are 
not permanent but temporary, owing their origin to the various nutri- 
tive conditions afforded by different hosts. 
I have already called attention to the fact that some specialized 
races are distinguished by their ability to infect some hosts more fully 
than others. The races of Puccinia Ribesu-Caricis, P. Hieracu, P. 
graminis, Uromyces Dactylidis and Coleosporium Campanulae afford 
illustrations of this condition. The infection or non-infection of 
certain hosts is dependent to a great extent upon particularly favorable 
experimental conditions. It is well known that various factors do 
