SEEPAGE OK RETURN WATERS FROM IRRIGATION. 
53 
found in 1895 is probably due to this cause. The inflow for 1895 
was sixty second-feet more than the average. The rainfall was over 
four inches more than the average. Yet the extra sixty feet 
throughout the year would be given by a depth of one inch over 
40,000 acres. If this comes from the rainfall, we must conclude 
that but very little of the extra rainfall was effective. As irrigation 
water is applied more freely because the supply in the river is 
greater, it seems more probable that the larger amount is due rather 
to the more water used than directly to the greater rainfall, though 
at present the effects cannot be entirely separated. 
§ 50. Direct evidence bearing on the question was sought in 
the Platte measurement of 1894, but with negative results. If 
there be any substantial increase from such source, then the chan¬ 
nels which conduct the drainage from a large area should show 
some indications of it. There are a number of such channels 
leading into the Platte, each of which drains over 1,000 square 
miles. This is more than the mountain water-shed of the Poudre 
river above its exit from the mountains. As the surface of the 
rock or impermeable surface has the same undulations as the sur¬ 
face of the ground, the underground drainage must follow essen¬ 
tially the same lines of drainage as the surface. This is shown 
plainly in the excellent sections taken at various points across the 
plains by Col. Nettleton and Mr. Follett. * One of these sections 
was across the valley of the Platte at Sterling. 
The streams following these drainage lines, while permanent 
near their upper ends, are almost never flowing near their outlets 
into the Platte. It has generally been believed that these streams 
furnish much water to the Platte through the sand of their beds, 
and it has been a favorite article of belief among the adherents in 
the underflow idea. If this be the case, it ought to be shown by 
taking a measurement of the river above the mouth of the stream 
and below, far enough apart to include the bed of sand forming the 
channel. Even if the increase is not noticeably great, the rate of 
increase might well be expected to be greater than for the average 
of the stream. 
§ 51. In order to test the question, I instructed the observers, 
in 1894, to measure the river above the important drainage chan¬ 
nels, and also below. This was done by Messrs. Trimble and 
Preston, with the results shown in the detailed tables, and brought 
together in Table X. In most cases the channel spreads out into 
the bottoms of the Platte, so that it is sometimes necessary to make 
the measurements several miles apart, in order to include the ex¬ 
pected inflow. 
* Reports Artesian and Underflow Investigation, 1890-1,1892, U. S. Depart¬ 
ment Agriculture. 
