40 Prof. J. H. Poynting on 



applicable to both solids and liquids, and obtains a more 

 general form for the result. 



Let a volume v of a body (solid or liquid) be in a perfectly 

 conducting cylinder (fig. 1, Plate L) so arranged that the tem- 

 perature is always constant. A porous plug, which the sub- 

 stance if liquid does not wet, is in the cylinder ; and the holes 

 in the plug are to be so fine that any required pressure can be 

 applied to the liquid without forcing it beyond the further sur- 

 face of the plug, the curved surface of the liquid there with- 

 standing the pressure. A piston to which pressure can be 

 applied is in contact with the substance : and beyond the plug 

 is another movable piston to which any pressure can be applied, 

 the arrows in the figure indicating the direction in which the 

 external pressures are applied to the pistons. 



Let the volume of the substance in the denser state at the 

 pressure of its normal vapour-tension trfor the given tempera- 

 ture be v. Let V be the volume of the whole as vapour at the 

 pressure nr. Let the equilibrium vapour-tension when the 

 denser state is subjected to a greater pressure ot + P he p, and 

 let the volume of the whole as vapour at the pressure p be 

 V. Let the coefficient of cubic compressibility be k. Now 

 take the body through the following cycle. 



Increase the pressure to ts + P on the left-hand piston, and 

 then let the substance evaporate through the plug to the right 

 hand, pushing out the piston there at pressure p till the whole 

 is evaporated to a volume V'. lip be greater than -sr, let the 

 vapour expand from V 7 , always in equilibrium with the pres- 

 sure, finally arriving at a volume V and pressure w. Now 

 cover the porous plug, and, if necessary, commence condensa- 

 tion by introducing a small amount of the substance. Push 

 in the right-hand piston at the pressure is till the whole is 

 condensed to volume v. 



We have now conducted the substance through a cycle each 

 step of which is reversible*. Then 



V/Q 



$' 



T -o. 



* It seems difficult to imagine a plug which would satisfy the condition 

 of reversibility for the solid under great pressure in contact with its vapour. 

 Perhaps the following would answer the requirements, if an ordinary 

 porous plug is insufficient. Suppose the solid in a finely-divided state, 

 and contained in a liquid which wets it but is of a very slightly greater 

 specific gravity, and whose vapour-tension is negligible. During evapo- 

 ration turn the cylinder with the vapour-chamber upwards. The particles 

 of solid will rise up through the pores, and a small fraction of their surface 

 will protrude, but they will otherwise be subjected to a pressure nr+p. 

 For condensation reverse the cylinder. As the solid condenses on the 

 surface it will rise up as fast as it is formed, and so increase the volume 

 of the chamber and force back the piston, 



