COLORADO IRRIGATION WATERS AND THEIR CHANGES. 75 
47. The amount of potash salts removed by the drain water is less 
than that existing in solution in the ground water. The total amount 
removed in the course of years is a large one and, while we are re¬ 
minded that the draining is going on all the time, day and night, year 
after, year, we have to consider also that potash is not taken from any 
single acre-foot of soil, nor from a mass represented by a single acre of 
surface, but for the sake of keeping our numbers within limits which we 
can appreciate, I will give the figures showing the amounts of potash, 
on the basis of the acre-foot. One acre-foot of our soil contains a total 
of 78,750 pounds. Of this dilute hydrochloric acid dissolves 43,750 
pounds. An acre-foot of ground water, before irrigation, in 1898, con¬ 
tained 22 pounds, and in 1899, 6 pounds of potash. After irrigation, in 
1898, it contained 41 pounds, in 1899, 18 pounds. An acre-foot of drain 
water carries but 5 pounds, taking the average of four drains. The 
water draining from any given acre of land is probably small, not ex¬ 
ceeding a hundredth of an acre-foot daily, in which case the amount re¬ 
moved from any single acre of land is very small. We will put this 
another way, in which the statements may seem more definite. If we 
take a strip of our land 18 miles long and one mile wide, from which 
there is discharged 3t) acre-feet of drainage water daily, it would take 
upwards of 50 ,<>00 years for it to carry out an amount of potash equal to 
that contained in the first three feet of soil. 
48. The nitrogen carried by the drain waters is only a little more 
per acre-foot than the potash, it being 5.8 pounds. The supply of nitro¬ 
gen in the soil is not so great as that of the potash, by any means, but 
while there is no accession of potash, except it be added, there may be 
of nitrogen. The amount found in the first two acre-feet of our soil was 
7,000 pounds, and it would take 1,227 acre-feet of drain water to contain 
this much nitrogen. 
49. The return waters furnish us a bigger and slightly different 
measure for the effects of drainage, and as this, with us, is mainly due 
to irrigation, they furnish us, perhaps, the best criteria by which to 
judge of the effects of irrigation upon our lands. We would expect to 
find their composition very similar to that of the drain waters, provided 
our samples of drain water were numerous enough to represent the va¬ 
rious soils and conditions to be met with in the 176,848 acres of land 
nominally under irrigation within this valley, of which probably 140,000 
are actually irrigated. This remark applies, of course, to the Poudre 
Valley and river. 
50. We find the total solids in the return waters lower than in the 
ground waters, and having the same range as found for the drain wa¬ 
ters. We find them characterized by the same salts, and in the same 
order in regard to their relative quantities; i.‘ e., calcic sulfate, magnesic 
sulfate and sodic carbonate, with sodic sulfate irregular in its quantity, 
but always subordinate, except in the sample of Arkansas river water, 
taken at Rocky Ford, April 24, 1903, concerning which some doubts may 
be entertained, but which is probably correct, because the ground wa¬ 
ters of that section are extremely rich in sodic sulfate. 
51. The salts discharged by the Poudre into the Platte river do not 
amount to less than 22,000 tons annually, which come from the lower 
section of the river, besides what may be carried from sections further 
up the river when the waters are not all taken out, as was the case at 
the time our samples were taken. 
52. The chief difference between the drain waters and the return 
waters taken from the rivers, is in the potash present, which is greater 
in the return waters than in the drain waters. While some of the drain 
waters contain almost as much potash as the return waters, the latter 
are, as a rule, richer in potash than the former. The main features of 
these two classes of waters are, however, identical. 
53. The waters of other streams, which we examined fully, justify 
us in assuming that the Poudre river water is typical of the mountain 
waters on this side of the range. They show that the waters of the dif¬ 
ferent streams, while in their mountain sections, are identical; that all 
