372 
Journal of Agricultural Research 
Vol. XXV, No. 9 
spring regardless of whether they were winter or spring wheats. Of 
course the winter varieties did not head when sown in the spring nursery, 
so that it was impossible to get notes on head infection. 
The methods of making inoculations in order to induce an epiphytotic 
of stripe rust in the nursery were such as seemed to offer the best condi¬ 
tions possible for an abundant and widespread infection. Border rows 
of Chul, one of the most susceptible varieties of wheat, were sown around 
all of the plots. These border rows were inoculated both by spraying a 
suspension of urediniospores over their entire length and by hand inocu¬ 
lation of individual plants at intervals in them. 
In addition to the border-row inoculations, one end of each varietal 
row was sprayed with a spore decoction and individual plants in every 
row were hand inoculated. Hand inoculations were made by smearing 
spores on the leaves with a scalpel, spraying the plants with water, and 
covering the inoculated plants with inverted flower pots for a period of 
48 hours. The quantity of infection on the border row served as a control, 
showing whether or not climatic and other conditions were favorable for 
abundant infection. 
The results of the 3-year study of the effects of such inoculations on 
163 varieties of'common, club, poulard, durum, and Polish wheats, and 
eminer, spelt, and einkom, when grown in the nursery, are shown in 
Tables III and IV. 
The method used to indicate the quantity of infection recorded in 
Table IV needs a word of explanation. It seemed to the writers that 
the usual method of estimating the quantity of rust on cereals was not 
the best for accurately indicating the quantity of stripe rust present. It 
was found that varieties differed widely in the proportion of the plants 
in the row which showed infection, as well as in the quantity of rust on 
those plants which were rusted. In other words, there was much varia¬ 
tion in the rapidity and extent of the spread of the rust on the different 
varieties. 
Table III shows, for each variety, the proportion of the plants which 
were rusted at the Oregon Agricultural Experiment Station in the fall 
of 1917 and the spring of 1918, expressed in percentages, and the degree 
of rustiness of these infected plants, expressed in terms of a scale ranging 
from 1 to 10. The degree of infection refers to the proportion of plant 
surface covered with rust. This necessitates the recording of two 
numbers to indicate the comparative susceptibility of each variety. 
In order to reduce these figures to a single product which would express 
at a glance this comparative susceptibility, the following plan was 
devised. The product of the percentage of plants infected by the degree 
of infection would give in comparative terms the average infection of 
any given variety, with a possible maximum of 1,000, where 100 per 
cent of the plants in the row were infected with a maximum degree of 10. 
This method of computation has been used to indicate infection of the 
plants when in leaf and when heading. Infection data were taken 
when about 5 leaves were out, and again when the heads were well 
out of the boot. 
As an example of the application of this method, take from Table 
III the data for Chul when grown as a winter wheat in 1917-18. The 
percentages of infected plants recorded for the first, second, and third 
replications are 80,100, and 70, respectively, and the degrees of infection 
are 6, 6, and 3, respectively. The sum of the products of each per- 
