1 2 Journal of Agricultural Research voi. v, no. i 
Five determinations were made of the plants in each glass, making about 
18,450 separate determinations. This number was reduced somewhat 
by the fact that plants did not germinate in all the glasses, owing to the 
high salt content. With this great number of results it is impracticable 
to give all the data in detail; hence, only summaries will be presented. 
COMBINATION OT SALTS 
In each test containing 15 glasses three different salts were used. 
The glasses were arranged in the triangular diagram used in expressing 
SODIUM CHLORID 0- SODIUM SULPHATE ®-SODIUM CARBONATE 
z - Diagram showing percentage of salts, mixtures, and their position in the diagrams of experimental 
sets. The arrangement of salts shown here is that in figure 2, page 14 . The same positions but with 
different salts apply to figures 2 to 25. 
three variables. This arrangement is shown in figure i, the salts in 
this case being sodium sulphate, sodium carbonate, and sodium chlorid. 
All 15 glasses contain the same total concentration of salts—for example, 
in figure i the concentration is i ,000 parts of salt per million parts of 
dry soil. 
The glasses on the corners of the diagram which are marked “A,” 41 E,” 
and “O” contained 100 per cent of the single salts. The other glasses 
