CLXXXII. GAUGING OF FLOW OF STREAMS AND ARTESIAN BORES. 
but has found that the slope measured over 2,000 feet 
differed from that over 1,000 feet, and that the slope at 
one side differed from that at the other side. 
In practice the slope must be measured over one or two 
thousand feet, with a level and staff, and it is evident that 
the slope thus found is not the ‘s’ in the formula, so it must 
be assumed that this is equal to the local slope or bears 
some definite relationship to it, which is not correct except 
by accident. 
Starling says,’ ‘‘Hven the local slope is not constant, but 
varies greatly from bank to bank.’’ TT. G. Ellis found that 
by measuring the slope carefully over 100 and 400 feet 
lengths, and using the best modern formule, the discharge 
so found differed by 50 to 250 per cent. from that found by 
actual gauging. He thinks the slope is so uncertain an 
element that slope formule are of little value.” Cunningham 
came to the same conclusion in experiments on the Ganges 
Oanal.’ Hamilton Smith Jnr., says,* “All determinations 
of the flow in large streams depending on the factors s, R, 
and v are necessarily very unreliable.”’ 
Rod Floats.—This is the best of all float determinations 
of velocity, and was practically the only method used by 
the department till the beginning of 1902. The first 
authentic gaugings in New South Wales were made by Mr. 
H. G. McKinney, m. mst.o.m, in 1886, who adopted the rod 
floats in preference to the current meter, for the reason 
that the only meter at his disposal was one of the Revy 
pattern, and he states “‘In my time current meters were 
so unreliable that I did not care to go on using them; I 
came to the conclusion that the method I used (rod floats) 
was the better one. I tried a Revy meter in still water, 
— 
* Trans. Am. Soc. C.E., Vol. xxxiv., No. 5, p. 398. 
* Trans. Am. Soc. C.E., Vol. x1., p. 23. 
3 Proc. Inst. C.E., Vol. Lxxx1., p. 11. * Hydraulics, p. 191. 
