June i6,1923 Azotohacier Flora and Nitrogen Fixing Ability of Soils 909 
In this examination of soils for Azotobacter it will be noted that three 
methods of detecting their presence were employed—^the formation of a 
film, the microscopic examination, and the quantity of nitrogen fixed. 
Under the heading “Azotobacter” in Tables V, VII, IX, and XI a + 
(plus) sign has been placed where, in the opinion of the writer, the major 
evidence indicated the existence of a vigorous Azotobacter flora in the 
original sample of soil and a — (minus) sign where the evidence did not 
indicate the presence of a vigorous Azotobacter flora. 
The colorimetric hydrogen-ion determinations were made by the 
Clark and Tubs method as modified for soils by Gillespie. The soils were 
ground to pass a 40-mesh sieve and a weighed quantity mixed with five 
times its weight of water, shaken well and centrifugalized until the super¬ 
natant liquid was practically clear. The water used in the preparation of 
the soil extract and suspension was freshly distilled from a mixture of 
sulphuric acid and potassium dichromate into a flask containing barium 
hydroxid. From this it was distilled into a third empty flask and again 
distilled. The Ph of water thus obtained was from 5.7 to 6.0 and was 
affected by the minutest trace of acid or alkali. All glassware coming in 
contact with the soil extract was washed in this water. Buffer solutions 
were prepared according to Clark and Lubs (i) and were checked, and 
adjusted if necessary, at frequent intervals on a Leeds and Northrup type 
K potentiometer. 
Electrometric hydrogen-ion concentrations, or differences in potential 
between the soil solution and the hydrogen electrode, were made by 
using a Leeds and Northrup type K potentiometer in connection with 
saturated KCl —calomel and hydrogen electrodes similar to the one 
described by Hildebrand (9). The ratio of soil to water used was the 
same as employed in colorimetric determinations; i. e., i to 5. Six hydro¬ 
gen electrodes were connected by switches to the potentiometer so that 
six samples could be run at the same time. Hydrogen was bubbled 
through the cells continuously at a rather rapid rate, the cells being con¬ 
stantly shaken. A maximum difference in potential was usually recorded 
in 10 to 30 minutes, after which the difference decreased very slowly. The 
length of time required, to reach the maximum reading apparently depended, 
other things being equal, upon the rate of flow of hydrogen. The influence 
of the rate of flow of hydrogen upon the length of time necessary to obtain 
maximum difference of potential is illustrated in the data presented in 
Tables I and II. Neutral or alkaline soils usually required a longer time 
to reach the maximum difference in potential, and the agreement between 
duplicates was not, as a rule, as close as it was with acid soils. 
Tabi,^ I .—Time required for electrodes to record maximum difference in potential; hydro¬ 
gen parsed over electrodes slowly {readings recorded as millivolts) 
376. 
376. 
380. 
380. 
381. 
381. 
410. 
410. 
Soil No, 
IS 20 
613 
618 
503 
517 
Time (in minutes). 
520 
300 
633 
654 
623 
6 S 4 
51S 
54 ° 
35 
40 
523 
524 
642 
691 
45 
SO 
655 
678 
563 
^566 
SS 4 
563 
“524 
<^526 
55 
565 
566 
524 
526 
670 
697 
60 
70 
80 
90 
95 
100 
673 
685 
^^566 
566 
683 
690 
687 
694 
690 
69s 
.... 
^693 
^696 
— 
695 
a'jo 2 
^703 
702 
703 
700 
O' Maximum reading. 
