78 
Journal of Agricultural Research 
Vol. II, No. 2 
This station consists of a car, measuring accessories, and a concrete 
trough with a level bottom. The trough is 2 feet wide on the bottom 
and 3 feet deep, with side slopes of 1 to 1. At the time the experiments 
were made it was carefully cleaned of silt and debris and filled about 
three-fourths full with fresh ditch water carrying a rather high percentage 
of silt. 
The experiments at Calexico consisted of runs with the meter on a rod 
and held horizontal at a depth of 1 foot below the surface of the water, 
practically reproducing the conditions under which it was rated, runs 
with the meter tipped upward and downward at various angles, runs 
with the pivot dulled, runs with the meter held close to the side walls of 
the trough, and runs with the meter just clearing the bottom of the 
trough. The results of these experiments, plotted to logarithmic scale, 
are shown in figure 1. 
The meter used was rated at Chevy Chase Lake, near Washington, 
D. C., during the first half of May, 1913, by the United States Bureau of 
Standards and had been in use throughout the summer. The points 
developed by this original rating for the runs made with the meter 
equipped with a single-point contact head, as it was during the experi¬ 
ments at Calexico, are also shown in figure 1, for comparison with the 
Calexico results. The curve shown on the figure represents the results 
of the original rating. 
For the meter submerged 1 foot, the rod held vertical, and with the 
pivot in good condition the points are as close to the curve of the original 
rating as a majority of the points in the original rating, but are all on 
one side of the curve, indicating that the meter was slightly faster after 
several months’ use than at first, but the difference is so slight as to be 
negligible. 
The runs made with the meter tipped out of the horizontal position 
indicate that the meter does not run true to the standard rating curve. 
Not enough runs were made under these conditions to develop curves, 
but the results indicate clearly that it is very important to keep the meter 
horizontal. 
The runs with the meter held close to the sloping sides of the trough 
gave points on both sides of the standard rating curve, but indicate that 
the meter runs true to the rating curve under these conditions. This is 
shown also by experiments at Cornell University, which will be discussed 
later. 
The runs with the meter just clearing the bottom of the concrete 
fflannel show that a correction is necessary when the meter is so used in 
practice. When plotted on a natural scale, the data shown in figure 1 
indicate that, with the meter used, the water would actually flow about 
0.06 foot per second faster than indicated by the velocity of the meter. 
On a natural scale the curve for the meter held at the bottom was quite 
parallel to the standard curve and 0.06 foot per second slower. 
