413 
of Edinburgh, Session 1879 - 80 . 
return current was also near the surface, as the water was not heated 
to its maximum density it was unable to sink into the colder water 
at the bottom, and so flowed back over it and under the surface 
current. While this was taking place the upper thermometer was 
rising rapidly, while the lower ones did not show any signs of rising 
for some time. As the temperature of the surface water increased 
the return current got deeper and deeper, till it flowed at the bottom. 
The bottom thermometer then began to rise. This current continued 
to flow till the bottom water had acquired the temperature of maxi- 
mum density, being some degrees below it previously. After all 
the water had acquired nearly the temperature of maximum density, 
a change took place. A new current began to flow on the surface 
from the heat, and over the previous surface current. The lower 
currents gradually became slower and slower, and at last stopped. 
After a time the surface current also ceased and all was still, the hot 
water resting quietly on the cold. 
It is quite possible that the dipping of the isothermal of 39 *2° 
referred to by Mr Buchanan, might give rise to a circulation of the 
water, but the conditions necessary for it to do so are not likely to 
happen, owing to the bad conducting power of water. The isother- 
mal of 39 ’2° would require to dip to a great amount, to a number of 
feet lower at the sides than in the middle of the lake , in order to give 
rise to a sufficient “head” to put the water in motion. And even 
when a circulation is so produced, the water flowing towards the cold 
region will not be drawn upwards from the bottom , but will flow in 
horizontally , and will be water of a lower temperature than 39 '2°; 
that is, the lower returning current will descend but little below the 
lowest point of the 39 ‘2° isothermal, and all water at a lower level 
than this isothermal will be at rest, and retain its temperature 
of maximum density. 
This point was illustrated in the experiment described. After 
ice had formed under the end of the refrigerator, and after the bottom 
water had acquired a temperature of about its maximum density, and 
was quite still, a current was noticed flowing slowly from the ice at 
the surface, and a return current immediately underneath ; but more 
than a half of the water in the very shallow trough was unaffected 
by it, showing that the dense water does not rise from the bottom 
to keep up such a current, but that its supply is drawn from the 
