REED: SPECIALIZATION OF PARASITIC FUNGI 357 
According to the various workers the specialized race Avenae 
occurs on a wide range of hosts. Eriksson (41) lists it on twenty 
species belonging to fourteen genera; Jaczewski (68) lists it on seven 
species belonging to six genera; Carleton (25) records it on nineteen 
species belonging to fifteen genera; Stakman and Piemeisel (149) 
record it on thirty-three species belonging to twenty-one genera. The 
specialized race Tritict is reported by Carleton on seventeen species 
belonging to seven genera; Stakman and Piemeisel list it on thirty- 
three species belonging to nine genera; Jaczewski lists it on six species 
belonging to five genera; Eriksson records it on the four cereals and 
states that this race seems less sharply fixed in its host relations than 
any of the others. Stakman and Piemeisel report another race, 
Tritict compacti, from the Palouse wheat region of western United 
States. This race is reported on twenty species belonging to six 
genera. The specialized race Secalis also occurs on a number of hosts, 
Eriksson reporting it on eleven species belonging to five genera, 
Jaczewski reporting it on six species belonging to four genera and 
Stakman and Piemeisel recording it on twenty-three species belonging 
to nine genera. According to Eriksson, Jaczewski and Carleton the 
specialized race Agrostis occurs only on species of Agrostis, while 
Stakman and Piemeisel report it on thirteen species of grasses be- 
longing to ten genera. The specialized race Arrhenatheri occurs on 
both Avena sativa and Arrhenatherum elatius. The races Airae, Poae, 
Calamagrostis and A perae are limited to species of a single genus. 
Eriksson and Henning (43) separated the timothy rust from the 
black-stem rust and made it a distinct species, Puccinia Phlei-pratensis. 
The separation was based largely on the fact that they were unable, 
with one doubtful exception, to infect the barberry with teleutospores 
from timothy. Other workers have made similar attempts but with- 
out success. Stakman and Piemeisel consider it best to regard timothy 
rust as a race of Puccinia graminis. 
Eriksson (34) reports successful inoculations with uredospores from 
timothy on Avena sativa, Festuca elatior and Secale cereale, negative 
results being obtained with Triticum vulgare and Hordeum vulgare. 
Johnson (69) transferred timothy rust, using uredospores, to Avena 
sativa, Arrhenatherum elatius, Dactylis glomerata, Festuca elatior, Poa 
compressa and Secale cereale. Inoculations on thirteen other grasses, 
including Triticum vulgare and Hordeum vulgare, gave negative results. 
Stakman and Piemeisel (149), however, were able to infect eighteen 
species of grasses belonging to thirteen genera, these including Avena 
sativa, Hordeum vulgare and Secale cereale, but not Triticum vulgare. 
They regard this race as being closely related to the race Avenae. 
It may also be noted that Eriksson lists barley (Hordeum vulgare) 
