10 
Proceedings of the Royal Society of Edinburgh. [Sess. 
To make the temperature seiche more evident, a layer of paraffin resting 
on a layer of fresh water was also used. Owing to the low density of 
the paraffin, a “ temperature seiche ” of greater amplitude could be produced, 
and the sharpness of the boundary between the paraffin and the water 
made the phenomenon more easily observed. 
Currents produced by Wind. 
Although the experiments which have been described were undertaken 
primarily with a view to demonstrating the nature of the temperature 
seiche, they directed my attention to the currents which take place in a 
loch as a result of the surface current directly produced by the wind. 
Surface water is driven along by the wind, and a slight transference is 
also due to the waves. The place of the water so transferred must be 
filled, and a “ return current ” is set up. Where the wind-current does not 
cover the whole length or breadth of the loch, this return current takes 
place largely on the surface, and the circulation produced is purely a 
surface circulation. But where, as in a narrow loch, the wind-current is 
over the whole surface, a return current takes place through the loch, and a 
vertical circulation takes place. 
From the observations made in the experimental trough, I incline to 
the view that, during the first period of the temperature cycle, when the loch 
is of uniform temperature, the return current is distributed all through the 
loch, and is consequently very slow. In the experimental trough, with a 
liquid of uniform density, colouring matter was introduced at the surface 
and the wind-current started. There followed a gradual mixing of the 
colouring matter all through the liquid, showing that in the experimental 
trough at least the effects of the currents produced by the wind were not 
confined to the surface. 
The conditions in an actual loch are very different, but there, too, any 
observations available go to show that the effects of wind at this period 
are felt to great depths. 
Fig. 5 shows the temperature distribution over part of Loch Ness 
on 17th October 1903, towards the end of the third period, and when there 
were no great differences of temperature. There had been a S.W. wind 
blowing for some days previous, which on the 17th October changed to a 
N.E. wind. This accounts for the direction of the upper isotherms. But 
the direction of the lower isotherms shows well the great depth to which 
the currents induced by wind are felt. Reference may also be made to 
