Sept, i, 1923 
Puccinia glumarum and Hosts for Variety iritici 365 
Since the completion of the foregoing studies, headquarters for the 
investigations on stripe rust have been transferred from Corvallis, Oreg., 
to Moscow, Idaho, and the research is now being carried on by the senior 
author and Mr. J. M. Raeder. Since the transfer, the following addi¬ 
tional hosts for Puccinia glumarum tritici have been determined by green¬ 
house inoculation: Agropyron tenerum longifolium Scribn. and Smith., A. 
acutum (DC.) Roem. and Schult., Bromus purgans latiglumis (Scribn.) 
Shear, B. lanuginosus Poir., B. erectus Huds., B. macrostachys Desf., B. 
rigidus Roth, B. frondosus (Shear) Woot. and Standi., B. richardsonii 
Link, B. adoensis Hochst., Hordeum maritimum Roth, H. bulbosum L., 
Phalaris canariensis L., and Sitanion hystrix (Nutt.) J. G. S. The 
grasses in this list, with those listed previously, comprise fifty-nine 
species of common wild grasses which are known to be hosts for stripe 
rust as it occurs in the United States. Thirty-three of these have been 
found naturally infected, and twenty-six of them determined by artificial 
inoculation with P. glumarum tritici. Doubtless other hosts will be 
added to this list as our knowledge of this rust becomes more complete. 
SPECIALIZED VARIETIES 
Eriksson (4) named five specialized varieties of Puccinia glumarum , 
based on inoculation experiments carried on by him. These were 
P. glumarum tritici on wheat, P. glumarum hordei on barley, P. glumarum 
secale on rye, P. glumarum elymi on Elymus arenarius and P. glumarum 
agropyri on A. repens . He states that the varieties on wheat and barley 
appear to be sharply fixed (scharf fixiert) ; that is, he could not secure 
infection upon barley or rye with the variety tritici , or upon wheat or 
rye with the variety hordei. The variety secale appeared to be less sharply 
fixed, since he secured slight infection upon wheat inoculated with this 
variety. He considered that he might have had a mixed culture in the 
latter case. Neither of the varieties on Agropyron repens and on Elymus 
arenarius would infect wheat, barley, or rye, nor was he able in either 
case to reinfect the original host. 
It appears then that Eriksson had good evidence for establishing the 
varieties P. glumarum tritici and P. glumarum hordei, but there is some 
question as to the evidence in the case of the other three varieties. 
In the field survey for stripe rust, conducted by the writers, an attempt 
has been made to record not only the name and location of infected 
hosts but also the name of other near-by grasses and cereals. It was 
thought that these notes might furnish valuable data regarding the 
spread of the rust from host to host in the field and also supplement the 
study on specialized varieties being conducted in the greenhouse. 
As stated above, Hordeum jubatum and Bromus marginatus are two of 
the most common hosts for P. glumarum in the northwestern States. 
Wherever these two grasses have been found growing near each other, if 
one was infected, the other usually was infected also. With few excep¬ 
tions, wherever infection was common on either of these two wild hosts, 
stripe rust could be found also on wheat if plants of a susceptible variety 
could be found growing near by. This also was true to a more limited 
extent of Elymus glaucus, Hordeum nodosum , Elymus canadensis , and 
Elymus condensates, although these latter grasses are much more limited 
in their distribution and do not appear to be as common hosts for stripe 
rust as are the two first named. 
