COLORADO IRRIGATION WATERS AND THEIR CHANGES. 5 1 
ceding the taking of the sample and they also apply to the sample 
taken from the drain east of the beet plot. 
• § 91. The sample taken from the drain underlying the beet 
plot in 1903 ought to be the nearest representative of the ground 
waters, analyses of which have been given. This drain was not laid 
at the time the samples of the ground water were taken. This 
sample ought to represent the drain water from this plot of ground. 
There had been but little or no rain for some time, the surface 
ground was frozen and the sample was taken on this date, Feb. 23, 
because we feared that a thaw might set in and we would have to 
wait a long time and perhaps never obtain a more representative 
drain water than that which we were then able to procure. The 
presence of strontic and lithic oxids in this analysis is what we 
would expect from what has been said in connection with the river 
water. They have been found present whenever tested for, but be¬ 
ing of subordinate importance they were not determined in the 
other samples, the one being included with the lime and the other 
with the sodic oxid. 
§ 92. The potassic oxid found in the ground waters varied 
from 0.01 of one per cent, of the total solids to 1.2 per cent., with 
an average of 0.262 for the 92 samples averaged; whereas the aver¬ 
age for the drain waters is 0.125 per cent., which calculated per 
acre-foot of water gives 20.7 pounds in the ground water to 5.0 
pounds in the drain water. From the point of its fertilizing 
value, this amount is not very significant, but it serves to show 
the ratio which exists between the amounts in the ground and 
drain waters or the extent to which the soil retains the potash, if 
we may put it that way. 
§ 93. In regard to the sodic salts we find a difference be¬ 
tween the sulfates and chlorids. Adopting the average percentage 
of sodic sulfate found in the total solids of well A in 1898, which 
is probably a little too high to be accurate but will represent 
the general facts with sufficient accuracy, we find in an acre- 
foot of the ground water 868 pounds of sodic sulfate, and in a like 
quantity of drain water 168 pounds, or one-fifth as much. In 
Bulletin No. 72 I called attention to the fact that the salts in 
solution fell as the water plane fell, the salts seeming to remain in 
the soil. I also called attention to the fact that the upper portions 
of the ground water were richer in total solids than the lower and 
at the same time contained higher percentages of sodic sulfate. I 
find in the drain water further proof of what I then observed by 
taking samples directly from the soil. We see that sodic sulfate 
does, not pass readily into the drain waters. Not only the absolute 
amount falls, but its relative amount, showing that the soil par¬ 
ticles retain it as there suggested. 
