May 25,1914 
Cup Current Meters 
83 
by the vertical movement on the basis of the standard rating (column 8). 
The two curves show a slight difference, probably due to two causes: 
First, the difference in depth of the water and the consequent difference 
in the relation between the number of turns at top and bottom and the 
distance through which the meter travels. Second, a difference in the 
manner in which different men operate the meter. For the higher rates 
of movement of the meter the points for the curve appear quite consistent. 
This is due to the fact that the turbine rotates all the time, even in still 
water, when the vertical movement is faster than about 0.2 foot per sec¬ 
ond. Below.that velocity of movement the turbine rotates part of the 
time and remains at rest part of the time. It was noticed that the tur¬ 
bine rotates part of a revolution when the motion was changed from up 
to down, and vice versa. At these points the turbine rotated a little 
and then came to rest again and remained so while the meter was being 
moved in a vertical direction to the next turning point. 
A general deduction from these curves is to the effect that all vertical 
movement of the meter tends to turn the wheel in the same direction as 
does the flowing water. For the slower movements of the meter this 
tendency is not sufficient to overcome the internal friction in the meter, 
but, when it is operated in running water, the water already has over¬ 
come this internal friction, and the force due to the vertical movement of 
the water is still available to turn the wheel and thus cause it to over- 
register. 
Again referring to figures 1 and 2, it appears that a cup meter under¬ 
registers at all points nearer the surface than about 0.3 foot and also 
while very near the bottom. Therefore, in a process of vertical integra¬ 
tion with a cup meter there is a tendency to overregistration at all times 
due to the vertical motion, and a tendency to underregistration while 
the meter is passing from about 0.3 foot below the surface up to the 
surface and back to that point and while near the bottom. These com¬ 
pensating tendencies account for the fact that if the meter is moved very 
uniformly and very slowly the integration method gives results as close 
to actual discharge as it does. 
