206 
Journal of Agricultural Research 
Vol. XXVI, No. 5 
Obviously, when variable temperatures are worked with, an infinite 
number of combinations may be employed. In this experiment it seemed 
wise to employ the simplest combination possible which would enable 
a comparison to be made between the disease-producing influence of 
controlled variable soil temperatures and the influence of a constant 
soil temperature equivalent to the mean of the variable. It was decided, 
therefore, to alternate the soil temperature as uniformly as possible 
between 14 0 and 30° C. once every 12 hours; that is, the soil was to 
reach the maximum of 30° during the afternoon (between 1 and 2 
o’clock) and to reach the minimum of 14 0 , 12 hours later (between 1 
and 2 a. m.). 
These temperatures were selected because they represent a reasonable 
soil fluctuation under field conditions, and because they lie on one side 
of the apex of the disease curve established by the “constant” soil 
temperature experiments with Harvest Queen wheat, as shown in 
figure 1. The particular time interval used was selected, not only on 
account of the fact that it conformed nearly to the condition in nature 
but because it divided the time between the upper and lower tempera¬ 
ture range into equal intervals. 
Three additional series were operated at constant temperatures of 
14 0 , 22 0 (mean of 14 0 and 30° C.), and 30°, in conjunction with the 
alternating (14 0 to 30°) series. 
The methods of conducting this experiment were the same as those 
used throughout the constant temperature series. One tank was devoted 
to each temperature and five pots were used in each tank, four of which 
contained the inoculated plants and one the uninoculated control plants. 
Harvest Queen wheat seed, Helminthosporium culture 392, and steri¬ 
lized loam soil containing 33 per cent of moisture, water free basis, were 
used. 
Soil temperatures in the constant series were controlled as described 
for the previous constant temperature experiments. In the case of the 
alternating temperature series control was obtained by means of a soil 
thermograph which was remodeled to serve both as a recording thermo¬ 
graph and a thermostat. By means of adjustable platinum points fixed 
to the inking arm and to the lever staff which is used to hold the inking 
arm away from the drum while changing records, it was possible to 
operate an electric spring switch and a water valve when the minimum 
or maximum temperatures were reached. 
For this experiment an electric heater was obtained which raised the 
temperature of the water from 14 0 to 30° C. in approximately 12 hours. 
In addition, a flow of cold water was passed through a swivel valve 
which was so regulated that it delivered sufficient water to lower the 
tank water temperature from 30° to 14 0 in approximately 12 hours. 
The heater and valve were then operated by an electric current controlled 
by the adjustable platinum contacts, set at 14 0 and 30°, on the record¬ 
ing soil thermograph. The control apparatus required setting after each 
operation. 
Owing to the slight irregularity in the water supply and to imperfec¬ 
tions in the control apparatus there were some slight variations in the 
soil temperature curves shown in figure 2, but in the main these curves 
seem fairly satisfactory and should justify consideration of the disease 
data obtained therefrom. 
