of Edinburgh, Session 1879 - 80 . 
411 
has attained a temperature of 39*2°, the warmer inflowing water no 
longer sinks, hut spreads itself over the colder water and keeps its 
place on the surface, and does not give rise to a vertical circulation, 
and carry its higher temperature down into the depths of the lake. 
As theory might not have taken into consideration all the condi- 
tions which exist in a lake while freezing, an experiment was made, 
in which the conditions of a freezing lake were imitated as closely as 
possible, to see if Mr Buchanan’s theory was correct, and to confirm or 
correct our previous knowledge. A trough 46 cm. long by 15 cm. 
broad and 12 cm. deep was filled with w r ater. In the trough were 
fixed three thermometers, one was placed 1 *5 cm. from the surface, 
one in the middle, and one 1 *5 cm. from the bottom. The trough 
was made with glass sides, so that by dropping some coloured solu- 
tion into the water, the direction, position, and velocity of the 
currents could be noted from time to time. Over the trough was 
placed a refrigerator, one end of which dipped into the trough and 
touched the water for a length of about 1 2 cm. This trough repre- 
sented in miniature a section of a lake in which might be observed 
all the changes in currents and temperatures which take place due 
to the heating, cooling, and freezing of the water. Observations of 
temperature and currents were made and recorded from time to time 
during twenty-two hours. All these observations were in keeping 
with the theoretical knowledge we at present possess, and did not 
give support to the theory that the cooling of the water below 39*2° 
could give rise to convection currents which could carry their lower 
temperature into the depths of the lake. 
The apparatus being arranged as described, the trough was filled 
with water at a temperature of 52-5° Fahr. A freezing mixture was 
put in the refrigerator ; when this was done, currents at once began 
to flow. The surface current flowing towards the cold, that is towards 
the end at which the refrigerator touched the water, and the return 
current flowing near the bottom. These currents kept moving for two 
hours, during which time all the thermometers fell at very nearly the 
same rate, and at the end of the two hours they all indicated a tem- 
perature of about 41 '5° The circulation during this time gradually 
got slower and slower, on account of the difference of density of the 
water on which the existence of these currents depended gradually 
diminishing. 
3 j) 
VOL. X. 
