360 $. H. BARRACLOUGH AND T. P. STRICKLAND. 
means of metal discs faced with rubber to prevent injury to the 
edges of the orifices. 
Table IT. 
oot /Approsimate tea oa 
Orifice of scale. 
em. em. 
i} 0°835 ‘90 
2 2°00 98 
3 3°22 1°04 
5. Entrance conditions.—The orifices discharged into a tin 
vessel with a wire gauze bottom, through which the water fell on 
to a piece of wood floating on the surface of the water in the 
back compartment of the inlet box, shewn in Fig. 1. From 
this compartment the water rose under the partition, and thus all 
disturbance due to influx was prevented. In the front of the box 
a triangular notch was cut to carry the channel, a watertight and 
yet flexible, joint between the two being obtained by means of 
sheet indiarubber. This joint allowed of the slope of the channel 
being varied through a wide range, without straining the channel 
in any way. To the top end of the channel and flush with its 
surface, was attached a piece of sheet tin ‘splayed out’ in the 
shape of a semi-conical funnel, through which the water entered 
the channel smoothly, no ripples appearing on its surface for @ 
distance of about two feet down the channel at the higher slopes, 
and for greater distances at the lower slopes. By means of bands 
of coloured liquids it could be seen that the flow was continuous 
at the top, no vortices being observable. 
6. The experimental channel.—The experimental channel was 
constructed of two carefully planed kauri pine planks 1} inch 
thick in the rough, neither of which showed the slightest defect 
in the way of cracks or knots. These planks were simply screwed 
together at right angles without any special form of joint, and, 
after the water had been flowing for a short time, proved to be 
perfectly water-tight.!. The channel was about six metres long 
1 The right angle section was selected for ge pointed out by Mr. 
Knibbs in his paper already referred to, p. 3 
