50 Preservation of Animal Food. 



the density of the air within the receiver. From the expan- 

 sion cylinder the cold expanded air is conveyed away by the 

 exhaust pipe to the room to be cooled. When used on board 

 ship, the hold should be divided into compartments, into 

 which must be packed the meat to be frozen ; it is not neces- 

 sary that the contents of one compartment should be frozen 

 at a time, as the air, though raised in temperature by con- 

 tact with the meat in the first compartment, would still be 

 colder than the air in the adjoining compartment. 



The temperature of air is raised 4° Fahr. for every inch of 

 mercurial pressure. Assuming that the pressure of one atmo- 

 sphere equals thirty inches of the mercurial column, we have 

 120° Fahr. of latent heat rendered sensible by the compres- 

 sion into half its space of any body of air. If we remove 

 this sensible 120°, and permit the compressed air to expand 

 under a constant pressure of one atmosphere, its temperature 

 will not be reduced .to the extent to which it was raised by 

 compression, that is to sety, it will not fall from 80° to — 40°. 

 This is obvious, since, under a constant pressure, the tem- 

 perature of any quantity of gas, to which heat cannot be 

 communicated, varies as its volume, and conversely. If the 

 compressed air in expanding were permitted (by removing 

 the constant pressure) to expand to its original volume, its 

 temperature would fall to — 40°, but, as the pressure is constant, 

 it cannot assume that volume and its relative temperature. 

 In order then to reduce the temperature of air at 80° to — 40°, 

 it would be necessary to render sensible more than 120° of 

 latent heat by compression. 



The experiments of M. Gay Lussac, since repeated by 

 MM. Rudberg, Magnus, Regnault, and Pouillet, have 

 afforded me the data for determining the volume of any gas at 

 any temperature. They show that in any gas or vapour, 

 under the same pressure, the increment of volume corre- 

 sponding to 1° is expressed by the 490th part of the volume 

 which it would have if reduced to the temperatui ; e of 32°. 

 This co-efficient of the expansion of gases and vapours is 

 not absolutely exact, since gases are subject to a small 

 difference in their rates of dilatation, and the dilatation of 

 the same gas is not absolutely the same at different pressures ; 

 but the inequality and variation are so small, that I have 

 not thought them of sufficient importance to be considered 

 elements in my investigations. I have assumed that all 

 gases and all vapours dilate uniformly, and in the same 

 degree as atmospheric air. 



