576 JAMES THOMSON, ESQ., ON THE EFFECT OF PRESSURE 
contact with the mass of water at 0°, which is proposed to be converted into ice, 
and let the piston be allowed to move back to the position it had at the com- 
mencement of the first process. During this second process, the temperature of 
the air would tend to sink on account of the expansion, but it is constrained to 
remain constant at 0° by the air being in communication with the freezing water, 
which cannot change its temperature so long as any of it remains unfrozen. 
Hence, so far as the air and the hand are concerned, this process has been exactly 
the converse of the former one. Thus the air has expanded through the same 
distance through which it was formerly compressed; and, since it has been con- 
stantly at the same temperature during both processes, the law of the variation 
of its pressure with its volume must have been the same in both. From this it 
follows, that the hand has received back exactly the same amount of mechanical 
work in the second process as it gave out in the first. By an analogous reason it 
is easily shewn, that the air also has received again exactly the same amount of 
heat as it gave out during its compression ; and, hence, it is now left in a condi- 
tion the same as that in which it was at the commencement of the first process. 
The only change which has been produced, then, is, that a certain quantity of heat 
has been abstracted from a small mass of water at 0°, and dispersed through an in- 
definite mass at the same temperature, the small mass having thus been converted into 
ice. This conclusion, it may be remarked, might be deduced at once by the appli- 
cation, to the freezing of water, of the general principle developed by Carnov, 
that no work is given out when heat passes from one body to another without a 
fall of temperature ; or rather by the application of the converse of this, which of 
course equally holds good, namely, that no work requires to be expended to make 
heat pass from one body to another at the same temperature. 
Next, to prove that the freezing point of water is lowered by an increase of 
the pressure to which the water is subjected :—Let a cylinder, of the same ima- 
ginary construction as that used in the foregoing demonstration, contain some 
air at 0° C. Let the bottom of the cylinder be placed in contact with the 
water of an indefinitely large lake, of which the temperature is above 0° by an 
infinitely small quantity; and let the air be subjected to compression by pressure 
applied by the hand to the piston. A certain amount of work is thus given from 
the hand to the air, and a certain amount of heat is given out from the air to the 
lake. Next, let the bottom of the cylinder be placed in communication with a 
small quantity of water at 0°, enclosed in a second imaginary cylinder similar in 
character to the first ; and let this water be, at the commencement, subject merely 
to the atmospheric pressure. Let, however, resistance be offered by the hand to 
any motion of the piston of this second cylinder which may take place. Things 
being in this state, let the piston of the cylinder containing the air move back to 
its original position. During this process part of the heat of the air becomes latent 
on account of the increase of volume. Thus the temperature of the air, from being 

