540 
AMERICAN JOURNAL OF BOTANY 
rvoi. io, 
the reaction were made, principally by the use of calcium carbonate and 
calcium oxide. The treatment of the soils was similar in all the experiments 
with the exception of the addition of the materials necessary to secure the 
desired reaction. Disregarding slight fluctuations in moisture and tempera¬ 
ture relations, the conditions for all the experiments were similar. 
Soil Reaction and its Determination 
It is well known that most of our common agricultural plants thrive 
best in soils well supplied with calcium carbonate. It has been very gener¬ 
ally assumed, therefore, that many plants require an alkaline reaction in 
the soil. Measurements of soil acidity have been designed for the purpose 
of determining the amount of lime required in order to bring about an 
alkaline condition, or to adjust the reaction to some arbitrarily selected 
end-point. A sufficiently accurate method of determining the end-point 
reached after the attempted adjustment has been lacking. Within the past 
few years, however, the hydrogen electrode has come to be used as an 
indicator of reaction in the soil. This method allows a determination of the 
hydrogen-ion concentration and affords an accurate measurement of the 
intensity of the acidity or alkalinity rather than the total acidity or alkalin¬ 
ity. A more rapid and a sufficiently accurate method for determining the 
hydrogen-ion concentration of soil extracts is the colorimetric method de¬ 
scribed by Clark and Lubs (2). Data obtained by this method agree very 
closely with those obtained by the electrometric method as determined by 
Gillespie (6) and Gillespie and Hurst (8). 
In the following experiments the determinations of the hydrogen-ion 
concentration of the soil extracts were made by the colorimetric method and 
expressed in the usual terms. This method has been fully described by 
Clark and Lubs (2) and need not be described again. It may be added, 
however, that the series of standard buffer solutions employed in this inves¬ 
tigation gave perfect agreement when compared with a corresponding series 
of standard buffer solutions (belonging to the Department of Bacteriology, 
University of Wisconsin), the hydrogen-ion concentration of which had been 
checked by the electrometric method. The method of obtaining the soil 
extracts was essentially that described by Gillespie (6). 
The air-dried soil was put through a fine sieve and fifteen grams were 
placed in a hard glass tube of 100 cc. capacity. The soil was treated with 
10 cc. of conductivity water and allowed to stand until thoroughly moist, 
which required about five minutes. Twenty cubic centimeters more of 
water was added, the tube was closed by the hand, shaken 100 times, and 
then allowed to stand for about five minutes while the filter was being pre¬ 
pared. Schleicher and Schull ashless filter paper was used. After having 
been shaken 100 times as before, the contents of the tube were poured upon 
the filter, and filtration was accomplished with moderate suction. With 
care in conducting the filtration process, extracts were obtained which in 
