24 
Colorado Experiment Station 
finished. The differences are much smaller than we would expect 
and will aggregate approximately one percent in favor of the sam¬ 
ples taken after irrigation. Some of the water had been removed 
by evaporation, but the greater portion of it sank rapidly to such 
a depth that evaporation took place slowly and from the ground 
surface only. An acre-foot of water is sufficient to increase the 
moisture in eleven acre-feet of soil by about six and one-quarter 
percent. We have no such increase. In the fourteen days elapsing 
since its application it has either evaporated from the surface or 
percolated through the soil to a greater depth than eleven feet, and 
this water moving through the soil has taken with it the greater 
part of the nitrates. In the samples taken 29 April, we find in the 
third foot 33.86 parts per million, and in the same foot of another 
sample taken at the same time we find 2.86 parts per million, but 
in the seventh, eighth and ninth foot vve find this reversed, the 
smaller quantities being in the first sample. The nitrates have been 
washed further down in the latter than in the former sample. 
The effect of the crop upon both the moisture and nitric nitro¬ 
gen in .the soil is evident from the samples taken 14 July, within 
the plots and in the spaces between the plots. Evidently the shad¬ 
ing of the ground causes the top six inches to show a slight excess 
over the surface of the unprotected ground, but the next six inches 
and the remaining three feet show differences in moisture up to five 
percent or rather more. The samples taken within the plots are low 
in nitric nitrogen, the maximum being 3.94 parts per million, while 
the corresponding sample from the space between the plots contained 
10.34 parts per million. In regard to the lower water content of the 
soil within the plot, no one will question its being due to the action 
of the plants and the difference represents principally the water 
taken up by the plants and given up to the air. Concerning the nitric 
nitrogen, two things may have happened. The occupancy of the 
ground by the plants may have prevented the formation of the ni¬ 
trates and there may never have been as large a quantity in the soil 
within the wheat plots as in the unoccupied space between the plots, 
or the plants may have used up the difference that we find between 
the nitric nitrogen content of the samples. We are sure that the lat¬ 
ter process took place, but we do not know certainly that the former 
supposition is not true. It is maintained that it is true. 
The differences in the total nitrogen present in the soil within 
the plots and in the spaces between the plots are so small, that we 
are inclined to attach no value to them, especially when we remem¬ 
ber that it is exceedingly difficult to select two samples of soil that 
are really comparable in this respect, and the difficulty of making 
chemical determinations which are free from all errors. Very slight 
