The; Fixation of Nitrogen. 91 
was gathered and put into a round bottomed bottle loosely closed 
with pieces of paste board and allowed to stand in the room where 
the sun shone on it part of the day. In this case we find a gain 
of 7.99 milligrams for each 100 grams of soil in 48 days, equal 
to 7.449 percent of the nitrogen originally present. In this con¬ 
nection an observation made in 1900 is interesting. The observa¬ 
tion is noted in Bulletin 65, p. 45, and raises this question of the 
increase of nitrogen in samples kept in the laboratory for some 
time, in the case mentioned, fifteen months, though air-dried, 
bottled and corked but not sealed. We found a gain of twenty- 
eight pounds of nitrogen per acre-foot, too small to be conclusive. 
Nitrification — The original sample contained nitric nitrogen 
at the beginning of the experiment equal to 0.0035 percent, and 
at the end of forty-eight days, the samples contained the following 
amounts: 
Percent nitric nitrogen in soil at Gain in percentage of nitric 
end of 4 8 days. nitrogen originally present 
in 48 days. 
1 
0.0060 
A 
2 
0.0060 
Average 0.00600 
71.43 
3 
0.0060 
1 
0.0060 
B 
2 
0.0060 
Average 0.00587 
67.71 
3 
0.0056 
1 
0.0080 
C 
2 
0.0070 
Average 0.00750 
114.24 
3 
Lost 
1 
0.0080 
D 
2 
0.0090 
Average 0.00833 
138.00 
3 
0.0080 
1 
0.0060 
E 
2 
0.0060 
Average 0.00600 
71.43 
3 
0.0060 
1 
0.0040 
F 
2 
0.0040 
Average 0.00417 
19.15 
3 
0.0045 
These tables show beyond a question that both fixation and 
nitrification have taken place, the former in far greater degree in 
absolute, but the latter in an even more marked degree in relative 
measures. I11 two samples the nitric nitrogen has much more than 
doubled. 
All of these samples except F had become decidedly darker 
on the surface than they were at the beginning. 
We have in these samples the reproduction of the main feat¬ 
ures presented in our fields, the production of nitrates, the fixation 
of nitrogen and the browning of the surface soil. The water con¬ 
tent of the samples was chosen at about 18.0 percent because my 
observations in the field have led me to the conclusion that the 
