918 
Journal of Agricultural Research 
Vol. XXXI, No. 10 
APPARATUS AND METHODS 
The apparatus used for maintaining the temperatures of the 
experiments herein reported was an adaptation of the Wisconsin tank 
method. Essentially, it consisted of a series of water baths for small 
soil cans. A large wooden bin about 4 by 20 feet was built in the 
greenhouse, and 12 round, galvanized tants, each 16 inches in diam¬ 
eter and 26 inches deep, were so arranged in the bin as to allow a 
maximum amount of space around each tank. The bottom of the 
bin and the spaces between the sides of it and the tanks were well 
packed with sawdust and straw as a means of partial insulation. A 
J^-inch brass tube was soldered into the side of each tank at the bot¬ 
tom, and a rubber tube attached which reached to the outside of the 
bin, where it was closed with a spring clamp. The bin and tanks were 
covered with boards closely fitted. Over each tank three holes 5 
inches in diameter were cut, the centers being equally spaced on a 
circle the radius of which was 3.5 inches, and the center oi which was 
approximately the center of the tank. At the center and between the 
three large holes a small hole was made to receive a thermometer. 
Thirty-six galvanized-iron cans, each 5 by 12 inches, were prepared 
with a wire lip at the top, so that they would not fall through the hole 
in the cover above the tank. 
The method of preparation of the soil and seed potatoes for use 
in these cans was simple. A uniform quantity of a suitable potato 
soil was prepared by much working and shoveling, and brought to a 
uniform moisture content of approximately 18 per cent. The soil 
and cans were sterilized by steaming. The soil was cooled and mixed 
again. Potatoes of the Early Ohio variety were used. They were 
of uniform size, each weighing about 4 ounces. The tubers were 
immersed in a 1 to 1,000 solution of mercuric chloride for one hour, 
then rinsed with sterile water. Following this, they were cut once 
lengthwise with a sterile knife, inoculated,‘and planted immediately. 
One can of the three for each tank was used as a control. The 
surface of the control seed piece was treated with 2 cubic centimeters 
of sterile distilled water. The seed pieces for infection were inocu¬ 
lated with 2 cubic centimeters of a heavy spore suspension of Fusarium 
oxysporum in sterile distilled water. Six seed pieces were planted in 
each can in two layers, three in each layer. The top layer came 
within 4 inches of the soil surface, and was about 5 inches below the 
surface of the water in the tank. The cans were brought to uniform 
weight by the addition of soil. They were then placed in the holes 
in the cover of the tanks. The tan£s had been previously filled to 
the top with w^ater and adjusted to the desired temperature. 
No automatic device was used to regulate or control the tempera¬ 
tures. The tanks were under observation continuously for the 
duration of the experiment. The thermometers used were carefully 
.calibrated, and corrections compensated for in the readings. The 
temperature of the water was raised by introducing an electric 
immersion heater, and lowered by the addition of cold water, with 
the removal of an equal volume at the bottom. Before and after 
each change the water was thoroughly agitated with a 4-inch propeller 
on a long shaft turned by a hand drill. In order to add water and 
insert the propeller a can had to be lifted. They were lifted in 
rotation. Thermometers in them detected no change of tempera- 
