440 Wisconsin Academy of Sciences , Arts, and Letters. 
and to include higher temperatures, the more elaborate plan which 
has since been followed was quickly evolved. It consists in having 
a series of water compartments or tanks in each of which the water 
can be held fairly constantly at any desired temperature. The 
Text Fig. 2. The second stage in the development of the soil-tempera¬ 
ture-control tanks. This was installed on a greenhouse bench over the 
regular steam heating pipes (G) which served to supplement other heat¬ 
ing devices such as live steam (F) and electric heating bulbs ( E). The 
lower temperatures were secured by cold running water. Insulation of 
the tank was obtained with hair felt (L) between the galvanized iron 
lining (M) and board (K), and an asbestos board cover (D). Tempera¬ 
tures from 5°C. to 40 °C. were obtainable and could be held fairly con¬ 
stant. In each compartment were sunk four battery jars filled with soil 
for the culture of the experimental plants. 
experimental plants are grown in glass or metal containers which 
can be sunk in these water tanks so as to maintain the soil 
temperature desired for the period of trial. To begin with, it 
seemed best to try to maintain the graduated temperatures by cir¬ 
culating the water through the compartments from the cooler to the 
warmer until the house temperature was reached. The development 
of this idea was undertaken by R. E. Hartman working in associa¬ 
tion with James Johnson on the tobacco root-rot problem. The 
compartment tank evolved, as illustrated in text figure 2, served its 
purpose well and has continued in use since 1915. It soon became 
