222 
Vol. XXIX, No. 5 
Journal of Agricultural Research 
plant. Aside from the teliospore ger¬ 
mination studies conducted by Eriks¬ 
son and Henning (4) in the early 90’s, 
but little further work has been done 
to determine either longevity or pos¬ 
sible function of the teliospores of this 
rust. The investigation of this phase 
of the stripe-rust problem by one of the 
writers (Hungerford) is now in progress 
at Moscow, Idaho. 
[sources of natural infection 
Until the two recognized races of 
Puccinia glumarum in North America 
have been critically studied in the light 
of their differential host relationships, 
knowledge of possible sources of natural 
infection will be but fragmentary and 
more or less uncertain. According to 
Hungerford (11), stripe rust already 
has been found to occur naturally in 
this country on wheat, barley, and rye, 
and 39 wild grasses. Two grasses very 
commonly infected are Horde urn juba- 
tum and Bromus marginatus; and the 
form of the rust occurring on these 
hosts goes readily to wheat. Hordeum 
jubatum is common throughout the 
Northern States from Washington to 
Maine, yet the rust has not traveled 
eastward beyond 103° W. longitude. 
It probably is true, however, that the 
wild-grass hosts play an important role 
in the spread of stripe-rust inoculum in 
the form of urediniospores that have 
survived the winter and new uredinio¬ 
spores which have been produced by 
hibernated mycelium. Irrespective of 
any wild-grass hosts, there is always 
the possible winter survival of both 
urediniospores and mycelium in the in¬ 
fected tissues of the cultivated hosts, 
wherever winter cereals are grown. 
Volunteer wheat, barley, and rye, de¬ 
veloping as they usually do very early 
after the first late summer rains, fre¬ 
quently become infected and serve as 
sources of inoculum for the infection of 
the fall-seeded crop. 
Hungerford (10) conducted extensive 
studies to determine the possible re¬ 
lation of wheat kernels infected with 
Puccinia graminis tritici to the occur¬ 
rence of initial infection in seedlings 
grown from such kernels. His evi¬ 
dence was wholly negative. The same 
author (11) obtained like results in his 
investigation of the problem of seed 
transmission of living mycelium and 
urediniospores of P. glumarum. 
Eriksson and Henning (4) assert that 
it may be accepted as an established 
fact that this rust frequently lives from 
one crop season to the next in a 
dormant or resting-mycelium stage, 
present in infected seedlings of winter 
wheat. They do not show, however, 
that such winter survival of the 
fungus is necessary to its perpetuation, 
even in the far northern agricultural 
districts of Sweden. Indeed, their 
experimental study of the relation of 
such factors as humidity, light, and 
temperature to germination of the 
urediniospores would indicate that 
low temperatures enhance rather than 
inhibit their germination. They do 
not deny the fact that hibernating 
urediniospores, under favorable condi¬ 
tions, may assist materially in accel¬ 
erating a severe epidemic of rust; but 
it is their opinion that the severity 
and extent of such an epidemic are 
more directly dependent on dormant 
mycelium. 
Biffen (2) states that stripe rust is 
the most destructive of the wheat rusts 
common to England and that winter 
survival of P. glumarum under the 
conditions of England’s winter climate 
is very largely due to viable uredinio¬ 
spores which have survived the winter. 
Mehta (19), in his comparative study 
of the viability and germination of the 
spores of cereal rusts, found the 
urediniospores of P. glumarum to be 
short-lived and their vitality impaired 
by temperatures easily withstood by 
the urediniospores of P. graminis and 
P. triticina. His experiments indicated 
that the urediniospores of P. glumarum , 
exposed to temperatures as low as 
2.5° to 5° C., retained their viability 
less tenaciously than do those of P. 
triticina. He found that 15 to 20 per 
cent of them survived an exposure of 
a month at the above temperatures. 
The writers’ observations on the 
probable overwintering of living 
mycelium agree with those of Eriksson 
and Henning. In connection with 
an experiment conducted by one of the 
writers (Hungerford), designed to deter¬ 
mine the possibility of seed transmis¬ 
sion of the disease, three lots of Chul 
wheat were sown in October, 1916. 
One lot of seed was badly infected with 
stripe rust. Another lot, similarly 
infected, had been given the modified 
hot-water treatment; and the third 
lot was of clean, rust-free seed. Some¬ 
what later, through a misunderstand¬ 
ing, some plants of Bromus marginatus 
and Hordeum jubatum infected with 
P. glumarum were set out adjacent to 
the above plats of wheat. During 
the winter, both grasses died down to 
the ground while the wheat remained 
green. On April 25, 1917, the rust 
was found on three or four wheat 
plants immediately next to these 
grasses. Subsequently, the rust spread 
rapidly from this center of infection 
