362 8. H. BARRACLOUGH AND T. P. STRICKLAND. 
was also levelled transversely, as nearly as possible, by wedges, but 
a perfectly accurate adjustment could not be obtained at every 
section and is evidently of small importance. At the lower end 
of the channel the water discharged without loss into an auxiliary 
channel through which it passed to a small vessel situated 
immediately above the gauging tank. From this vessel it could 
either flow directly into the gauging tank, or, by means of a shoot, 
to waste. 
7. Measurement of the quantity of water flowing through the 
channel.—The gauging of the water was done in a second 400 
gallon square iron tank, situated below the floor-line, as shown in 
the figure. To the sides of this tank also were bolted strong 
wooden stays so as to prevent any tendency towards bulging. The 
water was conveyed through a down pipe, from the small well on 
the top, almost to the bottom of the tank. A couple of air vents 
were left in the top of the tank. The water was discharged into 
a conveniently situated drain, through two valves fixed in the 
bottom. A gauge-glass and boxwood millimetre scale were 
attached to the side of the tank, in the same manner as already 
described for the supply tank. The cubic capacity of the tank 
throughout its entire depth was determined by means of a standard 
cubic foot, manufactured by Sugg for gas measurement. This 
was placed near the tank and connected with the water supply: 
It was alternately filled and discharged into the tank, a reading 
of the scale being taken before and after each cubic foot was run 
in. This work was done on perfectly calm days as it was found 
that any breeze produced slight oscillations of the water surface 
in the gauge-glass, and so prevented accurate readings being taken. 
To reduce cubic feet to cubic centimetres the multiplier 28,316 
was employed. The zero of the scale was several centimetres 
above the bottom of the tank, but this was of no importance as in 
any gauging it was the difference of two readings, and not 40 
absolute reading, that was used. Each cubic foot produced an 
average elevation of the water surface in the gauge-glass of 1-91 
