4 BULLETIN 832, U. §. DEPARTMENT OF AGRICULTURE. 
with a level to the nearest two or three thousandths of a foot and 
were carefully checked after every high stage in the stream to detect 
any possible changes due to the disturbance of the posts. The posts, 
however, were set in such a substantial manner, that very seldom 
were any changes found. The vertical measurements from the points 
on the arms to the water surface were made with a hook-gauge measur- 
ing rod. In some instances a stilling box arrangement was employed 
In connection with the hook gauge. These measurements were made 
either before or after the gauging was made, and often both before 
and after. They were made simultaneously by observers at either 
end of the course, or by one observer who first read at the end of the 
course nearer the gauging station, then at the other end, and then 
repeated the first measurement, the average of the first and third 
measurements being used. This method is based upon the assump- 
tion that the same time elapses between the first and second measure- 
ments as between the second and third measurements; hence errors 
due to a rising or failing stage are practically eliminated. Realizing 
that the accuracy of the results depended largely upon a correct 
determination of the slope, the utmost care was taken to secure accu- 
rate and dependable slope measurements. 
CROSS SECTIONS. 
The length of the slope course was carefully measured along one 
bank of the stream and stakes were placed at intervals of 50 or 100 
feet to mark the points where the cross sections were to be taken. 
Where the channel was regular the cross sections were taken at 
intervals of 100 feet. ‘Where the channel was irregular they were 
taken 50 feet apart, or at such irregular intervals as to represent as 
nearly as possible the true average cross section of the course. In 
making the cross-sectional measurements, distances across the channel 
were measured from the stakes to the nearest tenth of a foot. For 
the smaller channels, where wading was possible, elevations on the 
bottom and side slopes of the channels were taken with a level and 
rod to the nearest tenth of afoot, and were taken at intervals of 5 
feet across the channel and at ail points where a change in slope 
occurred. For the larger channels the measurements in the water 
section were generally taken from a boat attached to a rope stretched 
across the stream and fastened securely on both sides. 
COMPUTATIONS. 
In the experimental determination of the value of 2 in Kutter’s 
formula, 
1.811 . 0.00281 
41.6 vig re emmer 
. , 0.00281\ n_ 
frase 
(463 VR 
pth 
