12 
Proceedings of the Royal Society of Edinburgh. [Sess. 
end of the trough and rises to the surface. I endeavoured to obtain a 
photograph of this experiment, but was unsuccessful, as the return current 
very quickly mixed the colouring liquid all through the upper layer of 
water. 
By means of the colouring matter, it could also be observed at what 
point the return current began. There was always at the end of the trough 
towards which the wind was blowing a quantity of more or less passive 
water outside the ordinary current systems of the trough, and beyond this 
passive water the return current began. The point at which the return 
current began could be varied by removing part of the covering from the 
trough, and so making the wind-current less strong at the part of the trough 
which was uncovered. 
Fig. 6. 
But this return current at the surface of separation induced another 
return current in the lower layer. As it travelled along the surface of the 
brine, it carried some of the brine with it, exactly as the wind-current on 
the surface had carried some of the surface water with it. Consequently 
a secondary return current was set up in the lower layer, and in the same 
direction as the wind. This return current could be made evident experi- 
mentally by introducing a coloured liquid at the bottom of the trough, and 
the direction in which this liquid moved showed the direction of the secondary 
return current. Fig. 6 shows three photographs to illustrate the secondary 
return current. A layer of fresh water coloured black was introduced 
over uncoloured brine. Black stain was then introduced into the bottom 
of the trough by means of a pipette, and a current of air driven along the 
upper surface. The arrow on the photograph shows the direction of the 
