16 Colorado Experiment Station 
This section was taken to a depth of nine feet because at this 
point we struck coarse gravel and it was impossible to get the 
auger any deeper. 
We cannot infer that the water applied has washed out an 
amount of nitrates equivalent to the difference between the amounts 
shown by the first two and the last three samples, at least I do not 
consider it safe to make such a supposition, but four-foot samples 
taken subsequently will make it evident that large amounts of 
nitrates have actually been washed out of the soil. The amount of 
water used in this irrigation was one acre-foot, and the condition 
of the soil was such that this amount of water was necessary to 
permit the water, with the head that we had, to just flow over the 
soil. In other words, the ground took in this much water. The 
extent to which the irrigation had increased the water supply is 
fairly indicated by the differences in the percentages of moisture on 
29 April, when the first samples were taken and those found on 
27 June, 12 to 14 days after one acre-foot of water had been ap¬ 
plied. 
The nitric nitrogen in any foot of the last three samples is 
comparatively small; even the surface foot contains only a small 
amount, though it is considerably more than we find in any subse¬ 
quent foot. There was no nitric nitrogen in the last three feet of 
the first sample taken on 27 June, though this sample contained the 
largest amount in the surface foot at that time. 
The preceding samples were all taken in fallow strips, for we 
left spaces between each set of four plots. We did this so that we 
could take such samples without tramping down an unnecessarily 
large amount of our wheat. The loss of area due to taking of 
samples was too large as it was, for we also took samples in the 
wheat plots proper and cut many samples of plants. We, therefore, 
have two more series of samples to show the moisture, nitric nitro¬ 
gen and total nitrogen in the soil throughout the season. These 
two series were taken to a depth of four feet; one series was taken 
in the strips left between the plots and the other series in the plots 
themselves. Each of these series is divided into subseries corre¬ 
sponding with the three sections of land which we used for these 
field experiments. These sections of land are designated as 1700, 
1800 and 1900 and contain about two acres each. There were two 
sets of samples taken in each section on the dates given. These 
samples were not taken from one spot, but were composite. Each 
sample analyzed was composed of three samples which were well 
mixed before the final sample was taken out. These precautions 
were necessary in order to obtain a representative sample. I believe 
that the samples hereafter given are quite representative and de- 
