226 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9 
of the number of rust pustules on the upper surface of each leaf, of the 
length of each leaf in inches, of the extreme length of the entire plant, and 
of the number of stools it had produced. The plant was then dried, and 
its dry weight was obtained. The dry weight of the plant was adopted 
as the index of its relative vigor of growth, because more accurate sedation 
of the variables is possible on this value than on an index such as the 
height of the plant or the total leaf length. 
As an index of the degree of infection of the plant the value adopted 
was the number of pustules on an average unit area of the most severely 
infected leaf — calculated by dividing the number of rust pustules on the 
leaf by the length of the leaf in inches, and by its width at the base in six- 
teenths of an inch. This value was found to have a positive correlation 
(r = .7803 zt .0167 for the 250 variables of experiments I, II,, and III) 
with the value that at first thought would seem most desirable: namely, 
the total number of rust pustules counted on the leaves of the plant, divided 
by the total leaf length in inches, and by the largest leaf width in sixteenths 
of an inch — and is preferable for adoption in work of this kind not only 
because it is easier to obtain, but also because it avoids the error introduced 
by the development of new leaf surface during the incubation period of the 
rust. The most highly infected leaf on the plant was usually the lowest 
leaf in good condition. In tables 6-10 both values are given. 
Experiment I 
66 oat plants were grown in soil in 2-inch pots, divided into three groups 
on the basis of the number of plants grown to a pot. The soil was a rich 
garden loam. The seed was sown July 6, 1920, three grains being put 
into the soil for every plant desired, and the seedlings were later thinned out 
to the number of plants desired. The pots were kept on a bench in the 
greenhouse until August 11, when they were taken out of doors and set 
near a patch of rusty oats, subjecting the plants to natural conditions of 
inoculation and infection. The experiment was concluded on August 24. 
The data on this experiment are given in table 6. 
Table 6 
Group 
No. of 
Variables 
No. of 
Plants to a 
2-inch Pot 
Mean 
Dry Weight 
of Top of 
Plant 
in Mg. 
Mean No. of 
Pustules 
per Ave. Unit 
Area of Total 
Leaf Surface 
Mean No. of 
Pustules 
per Ave. Unit 
Area of most Severely 
Infected Leaf 
a 
23* 
5 
61 
I.I 
4.2 
b 
18 
2 
2.1 
6.0 
c 
25 
I 
183 
2.3 
6.1 
* Plus 2 destroyed. 
Experiment II 
70 plants were grown in 3-inch pots, divided into three groups on the basis 
of the number of plants grown to a pot. Soil, method of seeding, and dates 
