CONDENSATION AND CRITICAL PHENOMENA. 211 



the wave-shape of the curve gradually disappears, and 

 above a certain temperature has vanished altogether. The 

 point in which the wave disappears as it were has all the 

 properties of Andrews' critical point. In fact qualitatively 

 Andrews and Van der Waals' diagrams are exactly alike, at 

 least outside the border-curve. Inside the curve the theoret- 

 ical isothermal has the wave shape as described with points 

 between / and (j in which the condition of the substance is 



essentially unstable. 14 > J Experiment shows that 



the substance divides Itself into two portions, both stable 

 and also in equilibrium with each other, an equilibrium which 

 is independent of the qualities of the two " phases " present, 

 the pressure depending on the temperature only. The 

 kinetic theory is not able to determine, in a general way, 

 the condition of equilibrium, and in that way find points 

 belonging to the border-curve. Too little is known of the 

 nature of the superficial layer between liquid and vapour. 

 In cases of this kind we take refuge in thermodynamics. 

 The criterion to which it leads in this case was found by 

 Maxwell, and afterwards by Clausius, and may be stated 

 thus : the horizontal line which connects the two co-existing 

 phases on the isothermal encloses with the wave-crest and 

 the wave-trough surfaces of equal area. By this rule it is 

 possible to determine the vapour-pressure, and the vapour 

 and liquid density from the theoretical isothermal. But no 

 more than other rules arrived at by thermodynamical con- 

 siderations does it give us an insight into the mechanism on 

 which the equilibrium of the two phases depends. Kinet- 

 ically the condition of equilibrium is this : that the same 

 number of molecules passes from one phase into the other, 

 as backwards ; but, as was said before, this condition cannot 

 be applied unless by making special assumptions about the 

 nature of the bounding surface, which impairs the generality 

 of the result obtained. 



Quite lately, a most interesting theory was published 

 by Prof. Van der Waals regarding capillary phenomena, in 

 which he imagines the surface layer between the phases to 

 contain all the phases on the theoretical isothermal between 



