270 
Journal of Agricultural Research 
Vol. XIV, No. 1 
Very few, if any, local soils are regarded, agriculturally speaking, as 
deficient in lime. In fact, the application of lime far in excess of sup¬ 
posed requirements to soils on the Agronomy farm of the Kansas Experi¬ 
ment Station has been without effect upon productivity even when 
alfalfa was grown. Many of the soils herein reported as containing no 
Azotobacter were collected from the Agronomy farm. Sample 12 is 
from a plat to which lime has been applied, yet which failed to show 
Azotobacter. 
Since the writer was unable to associate the presence of Azotobacter 
with any other factor studied, and since the apparent-correlation of 
their presence with soil reaction was known, this factor was deemed 
worthy of investigation. A study of the influence of soil reaction 
seemed especially important, since more exact methods are now available 
for determining soil acidity. 
Recent investigation in other lines of bacteriology have shown that 
in many instances the degree of acidity or hydrogen-ion concentration 
is perhaps much more important in controlling bacterial activity than 
the total or titratable acidity. It has been shown in a number of 
instances that the degree of acidity tolerated by certain species of bac¬ 
teria has a very definite limit. Furthermore, none of the methods in 
vogue for ascertaining the total or titratable acidity of soils are very 
satisfactory. For these reasons it was thought best, if possible, to deter¬ 
mine the reaction in terms of hydrogen-ion concentration. For this 
purpose the writer has made use of the colorimetric method outlined by 
Clark and Lubs 1 as recently modified for soils by Gillespie. 3 The indi¬ 
cators used were methyl red, bromcresol purple, bromthymol blue, and 
phenol red. The standard hydrogen-ion-concentration solutions were 
prepared as directed by Clark and Tubs, and their accuracy tested and 
corrected, if need be, by means of electrometric measurements. Little 
difficulty was experienced in checking with different indicators except in 
those solutions falling on the acid side of bromcresol purple and the 
alkaline side of methyl red. Perhaps propyl red would have obviated 
this difficulty, but none was available when these analyses were made. 
However, any error arising from this difficulty can in no way vitiate 
conclusions that may be drawn from these experiments. 
In the last column of Table I is given the hydrogen-ion concentration 
observed in the soil extract, expressed in the usual way—that is, 
Ph. These results are certainly very striking. Of those soils in which 
no Azotobacter were observed, all with the exception of three gave a 
P H of 5.9 or less. All of the soils which gave Azotobacter growth, 
1 Clark, w. M., and Lubs, H. A. the colorimetric determination or hydrogen ion concen¬ 
tration AND ITS APPLICATIONS IN BACTERIOLOGY. In JoUT. Bact. ( V. 2, no. I, p. I—34; DO. 2, p. 109-136, 
no. 3. P* 191-236, 8 fig. 1917. References, p. 233-236. 
8 Gillespie, L. J. the reaction op soil and measurements op hydrogen*ion concentration. 
In Jour. Wash. Acad. Sd., v. 6, no. i, p. 7-16, 2 fig. 1916. 
