H. S. I. SMAIL. CLXXXIII, 
and found that it recorded so unsatisfactorily, that I decided 
that there was more objection to it than to the rods.’” 
Cunningham seemed to have a similar opinion about rod 
floats and current meters of the above type. 
In the author’s opinion the rod floats are better even in 
rivers with irregular beds, than meters of the Reévy 
pattern with recording dials under the water; this will be 
referred to later on under current meters. The method 
of procedure in gauging a river with rod floats is as follows: 
First of all the gauging site has to be selected; and the 
following applies to both rod floats and current meters. 
The ideal site is on a straight reach, far enough from a 
bend to be out of its influence; the bed should be per- 
manent and not stony. The banks should be sufficiently 
high to contain-all the water at the highest stages. The 
section should be free from all disturbing influences, such 
as bridge piers, etc. It should be such a place that the 
velocity is measurable even in the lowest stages, that is it 
should not be ina pool. As most of our gauges are bolted 
to bridge piers, gaugings cannot be taken at this section, so 
the site should be as near the gauge as possible. Seldom 
are all these advantages to be found at the one site, scour 
of bed or silting give trouble, at high stages the velocities 
are too great and ai low stages too small to measure 
accurately, so that the engineer must make the most of the 
best site he can find, always having in mind the desirable 
and undesirable qualities. 
When rod floats are being used, a search for a length of 
river with even cross section should be made, as a rule this 
will not be found, but often a place can be found in which 
a middle cross section is a mean between an upper and a 
lower one. Thus at Cowra at the gauging site for rod floats 
1 Minutes of Evidence Interstate Royal Commission on River Murray, 
p. 276. 
