The: Fixation of Nitrogejn. 31 
The soil designated as Laboratory No. 9§9ffi was very brown 
on its sui face, and when exposed to the atmosphere, after drying at 
ioo°, absorbed enough moisture to cause it to adhere so tena¬ 
ciously that one could press it into clumps in the hand which re¬ 
tained the form of the fingers. The aqueous extract had a brownish 
yellow color which became decidedly brown on evaporation. This 
color continued to go into solution even after the chlorids and 
sulfates had been completely washed out. The color could scarcely 
have been due to humus in the presence of such quantities of lime 
salts, besides the soil is poor in humus and the strong brown color 
on the surface of this soil is not due to this cause. It would have 
been difficult to have obtained a sample of this soil entirely free 
from the azotobacter pigments in which to determine the humus. 
Prof. Sackett has obtained cultures showing that the azotobacter 
pigments are soluble in water. This accounts for the deportment 
of this sample. 
The mechanical analysis given for Laboratory No. 1067 L 
quite as applicable to the sample No. 989 as to the one of which 
it was made and likewise the agricultural chemical analysis is 
quite as applicable to 1067. The localities are probably as much as 
two miles apart but the soils are similar in location and character. 
ANALYSIS 
Sand . 
Silicic acid (sol.) . . . . 
Sulfuric acid . 
Carbonic acid . 
Chlorin . 
Phosphoric acid . 
Lime . 
Magnesia . 
Potash . 
Soda. 
Ferric oxid . 
Aluminic oxid. 
Manganic oxid . 
Water at 100° C. 
Ignition . 
Sum . 
O. equivalent to chlorin 
XXXVIII 
Soil, top four 
inches 
laboratory 
No. 989 
Percent 
. 59.993 
. 15.000 
. 0.430 
. 3.450 
. 1.561 
. 0.220 
. 4.890 
. 1.783 
. 0.794 
. 1.557 
. 3.542 
. 2.189 
. 0.370 
. 0.500 
. (4.072) 
100.351 
0.351 
Total . 
Total nitrogen 
100.000 
0.118 
The ordinary agricultural analysis of this sample indicates 
that it is an excellent soil, which inference would be further 
justified by the mechanical analysis of No. 1067. While these 
inferences are fully justified by the analytical results and even 
further by the fact that all that was said of the preceding case, 
