4 
Colorado Experiment Station 
of fixation, for if the samples taken for this purpose at different dates 
had been taken on the same date, they would probably have varied in 
the same direction to such an extent as to render any conclusions based 
on them unsatisfactory, if not useless. It is not feasible to increase 
the number of subsamples sufficiently, in cases where one samples the 
land every fortnight, to eliminate this uncertainty. 
If it be objected that the fixation found in a laboratory experi¬ 
ment lasting from 25 to 30 days is not a safe basis for estimating the 
extent to which this process may take place in the soil, the answer is 
evident. The results of the laboratory experiments prove conclu¬ 
sively, especially where the experiment is made with the soil without 
addition of anything except water enough to produce a favorable de¬ 
gree of moisture, 18 to 20 percent, that the soil is populated by an 
efficient nitrogen-fixing flora. While a laboratory experiment extend¬ 
ing over 25 or 30 days does not prove that the fixation will proceed at 
the same rate for every other period of 25 or 30 days throughout the 
year, it does show that fixation is going on and may reach 105 parts 
per million, or 420 pounds of nitrogen for the surface acre foot, in 
one or more 30-day periods during the year. If we assume that the 
fixation is constant at one-half this rate, 50 parts per million, it would 
account for a large amount of nitrogen, the quivalent of a little more 
than 7 tons of proteid substance per annum. 
For the benefit of some reader not accustomed to think in these 
terms, it may be stated that a fifty-bushel crop of wheat will corres¬ 
pond to about 4,500 pounds of straw. If the wheat weighs 62 pounds 
per bushel and carries 2.0 per cent nitrogen while the straw carries 0.3 
percent nitrogen (this is a higher percentage than we have found in 
the stems and leaves taken together), the total nitrogen removed by 
the crop will amount to 74.0 pounds for each acre. If we assume the 
whole of this nitrogen to have been removed from the surface foot of 
soil, it is only three-fourths of the maximum fixation and one and one- 
half times the minimum obtained for a period of 30 days. The labora¬ 
tory experiments have shown that it is possible for this soil to obtain 
from the atmosphere as much nitrogen in 45 or 50 days as the assumed 
crop of 50 bushels per acre would remove. This process tends to keep 
np the supply of nitrogen, but it is almost certainly not uniform 
throughout the year. We cannot judge of the amount of fixation by 
the variation in the amount of nitrogen present, for there is a consid¬ 
erable wastage of nitrogen from our soils. That such wastage takes 
place is common knowledge, still we may illustrate it by the results of 
a single irrigation made in T 913 by which we moved about 300 pounds 
of nitrogen per acre beyond the reach of any ordinary, cultivated 
plant. This is only one cause of wastage. The writer does not know 
the results obtained by other observers, but his own results indicate 1 
a discrepancy of upwards of 300 pounds per acre for which he can> 
give no satisfactory accounting. 
