230 
Journal of Agricultural Research 
Vol. V, No. 6 
Table VI .—Discharge and velocity of various types of canals by measurements of two 
selected verticals 
Difference from correct discharge (per cent). 
Observations. 
Number 
of obser¬ 
vations. 
Average 
error of 
all obser¬ 
vations. 
Average 
variation 
Extent of variations. 
of a single 
observa¬ 
tion. 
Plus. 
Minus. 
Total discharge in flumes with vertical 
sides: 
For points one-fifth from sides. 
| 22 
/+ 1.7 
2. 1 
+ 4-5 
— 1.4 
For points one-sixth from sides. 
1.4 
2. 6 
+ 5-1 
— 6. 5 
Velocities only: 
Concrete-lined canal, steep side 
slopes— 
For points one-fifth from sides. 
} 10 
/“ 2.3 
2. 7 
+ 2. O 
- 5-7 
For points one-fourth from sides 
1+ 1.3 
3 * 1 
-f 6. 1 
- 3 - 7 
Concrete-lined canals, wide and 
flatly sloping sides— 
For points one-fifth from sides. 
1 « 
f- 1. O 
1. 9 
+ 1.6 
- 5*3 
For points one-fourth from sides 
/ 6 
\+ i- 4 
1.8 
+ 3-6 
— 1. 2 
Average earth canals, sloping 
sides— 
For points one-fifth from sides. 
■» 
/- 2 -3 
5 -S 
+ 9 - 0 
— 13.6 
For points one-fourth from sides 
1 + 4-3 
5 - 2 
+ 15-4 
- 3-8 
Average earth canals, steep sides— 
For points one-fifth from sides. 
| 20 
/- 4-2 
5-6 
+ 7.2 
-17.8 
For points one-fourth from sides. 
1 + 4-9 
5-7 
+ 12. 1 
~ 6. 7 
Earth canals, relatively deep sec¬ 
tions— 
For points one-fifth from sides. 
} 8 
/“ 5 - 7 
8.2 
-f* 10. 0 
— 19. 6 
For points one-fourth from sides 
1+ 2.6 
5-3 
— 9. 4 
+ I 3 -* 
Total discharges, points one-fifth from 
sides: 
Average earth canals, sloping sides.. 
15 1 
+ i-1 
5 *o 
+14.61 
-8.2 
Average earth canals, steep sides... 
20 
+ 3-3 
6. 0 
+ 14.2 j 
— 10. 4 
Earth canals, relatively deep. 
8 
+ 2. O 
7.0 
+ 14. 1 j 
—12. 1 
For vertical-sided flumes 22 gagings were available. The depths 
varied from 0.7 to 4.4 feet, the widths from 2 to 17.7 feet, and the dis¬ 
charges from 2 to 400 second-feet. The velocities and depths at points 
at a distance of one-fifth and one-sixth of the width from the sides were 
used to obtain discharges which were then compared with those ob¬ 
tained by the complete gaging. In such flumes with vertical sides the 
depths are practically uniform, and the use of the depth at only two points 
would cause little error in the resulting area. These results show that 
the two points whose mean velocities will equal that of the whole cross 
section lie generally between one-sixth and one-fifth of the width from 
the sides and that the error in using such index velocities at either pro¬ 
portion of the width averages about 2 per cent and does not exceed 5 
per cent, except in a few cases. 
For concrete-lined canals the canal section is uniform, and the cross- 
sectional area would be known for any depth. In such canals the dis¬ 
charge can be obtained from determinations of velocity and known areas 
for given depths. The comparisons given in Table VI are based on 
