96 TRANSACTIONS OF THE 
peated many times and the relative deviation of the meth- 
ods was determined. The methods used for determining 
the salts were as follows: First, the salt was added in sol- 
ution in various concentrations to the soil in tumblers. The 
soil ranged in this case from sand to clay including many 
mixtures with loam and varying organic materials. 
Second, at the conclusion of the experiment the soils in 
these tumblers were tested by water extraction for the 
salts added the method being described later. Third, 
these same soils were tested for total alkali with the 
electric bridge as described in U. 8. Bureau of Soils Bul- 
letin No. 61. Fourth, the soils were tested for total 
alkali by determining the freezing point of the soil at 
its normal moisture content, a method that has recently 
been perfected by Bouyoucos and McCool of the Michi- 
gan Station for determining the osmotic pressure and 
thus indirectly the concentration of the soil solution at 
the normal moisture content of the soil. 
The method of water extraction used was as follows: 
The soil was dried in an oven and a 50-gram sample 
weighed out. This was stirred in 500 cc. of distilled 
water with a broad paddle for five minutes, allowed to 
settle for an hour and filtered through a Chamberlain- 
Pasteur filter. The filtrate was tested for sodium chloride 
by titrating an aliquot part with N/50 silver nitrate using 
potassium dichromate as an indicator; it was tested for 
sodium carbonate by titrating an aliquot part with N/50 
sulfuric acid using methyl orange indicator and expressing 
the total basicity as sodium carbonate; it was tested for 
sodium sulfate by the ordinary gravimetric method. A 
rather extensive series of experiments showed that vary- 
ing the time of agitation and the proportion of soil to 
water had considerable influence on the sodium carbonate 
results but had little influence on the other salts. The 
method described above was adopted because it was con- 
venient and seemed to give as consistent results as any. 
The crop producing power of these soils was deter- 
mined by planting ten kernels of New Zealand wheat in 
each of the tumblers, maintaining the soil at a uniform, 
favorable moisture content and allowing the seed to ger- 
minate and grow for three weeks. The growth was then 
cut off at the surface of the soil, dried, and weighed. 
