24 
Steam Engine. 
the salt finely powdered, and of fresh -fallen and light snow.. 
On mixing these together, and immersing a thermometer 
in the mixture, the mercury sinks with great rapidity. For 
measuring exactly the cold produced, a spirit-thermome- 
ter, graduated to 50^ below 0 of Fahrenheit, or still lower, 
should be employed. A few pounds of the salt are suffi- 
cient to congeal a large mass of mercury. By means of 
13 pounds of the muriate, and an equal weight of snow, 
Messrs. Pepys and Allen froze 56 pounds of quicksilver 
into a solid mass. The mixture of the whole quantity of 
salt and snow, however, was not made at once, but part 
was expended in cooling the materials themselves. 
On a small scale it may be sufficient to employ two 
or three pounds of the salt. Let a few ounces of mercury, 
in a very thin glass retort, be immersed first in a mixture 
of one pound of each ; and when this has ceased to act, 
let another similar mixture be prepared. The second 
will never fail to congeal the quicksilver. 
The salt thus expended may be again evaporated, and 
crystallized for future experiments. 
The reader, who wishes for farther particulars respect- 
ing these experiments, is referred to the Philosophical 
Magazine, vol. hi. p. 76. 
VI. On the contrary^ liquids^ in becoming solid^ evolve 
or give out caloric ^ or^ m common language^ produce heat. 
1. Water, if kept perfectly free from agitation, may be 
cooled down several degrees below 32^ ; but, on shaking 
it, it immediately congeals, and the temperature rises to 
32 ^. 
2. Expose to the atmosphere, when at a temperature 
below freezing (for example, at 25^ of Fahrenheit), two 
equal quantities of water, in one only of which about a 
fourth of its weight of common salt has been dissolved 
The saline solution will be gradually cooled, without 
freezing, to 25^. The pure water will gradually descend 
