H. S I. SMAIL. CLXXXV. 
the boat about 30 or 40 feet above the upper section, so 
that it may acquire the full velocity before crossing the 
section. The exact time of the day when each rod passes 
beneath each wire is noted from the bank, and the distance 
on the wire noted from the boat. The author has found 
that an ordinary watch with seconds hand gives the best 
results, it can easily be read to half seconds. Then the 
time taken by the float to pass from the upper wire to the 
lower wire, into the distance between these wires, gives 
the mean velocity of the vertical, where the float cuts the 
middle wire, if the site has been judiciously chosen. 
When a river is rising or falling rapidly, it is impossible 
to measure three or more sections at each gauging, in the 
time at one’s disposal, although one can be easily be done, 
and if the middle cross section is practically the mean of 
all the cross sections, it may be taken as the cross section 
over which the discharge is measured. Ifa float strikes 
any obstacle on its run, such as a snag, etc., it is discarded, 
also if it takes a diagonal course through meeting eddies, 
not an uncommon occurrence even in straight reaches in 
made channels. 
The middle cross section is then divided up into several 
partial areas. let the width of each partial area be b 
feet. Then total discharge Q = b (vidi + vedo + ...... ) 
Where di, d., ds, etc., are the mean depths of each partial 
area, and V,, V2, Vs, etc, the measured mean velocity in 
each of these partial areas. The mean velocity of the whole 
section is then computed by dividing the total discharge by 
the total area. 
Advantages and disadvantages of this method. 
Advantages.—It is probably the very best method to use 
in an artificial channel of constant cross section, with float- 
ing grass and weeds, where a current meter would not be 
