CONDENSATION AND CRITICAL PHENOMENA. 205 



more complicated than vapour molecules (which, by the way, 

 is a very reasonable idea in itself) is also constantly used to 

 explain the abnormal phenomena near the critical point. 

 *' Normal " means here again agreeing with Andrews' dia- 

 gram. In heating a substance at constant volume (say in a 

 closed tube) the liquid surface ought to disappear only when 

 the volume is equal to the critical volume ; at greater 

 volumes the liquid ought to evaporate, at smaller volumes 

 the vapour ought to be entirely liquefied, in both cases 

 before the critical temperature is reached. Neither the 

 one nor the other is true ; the nunierous experiments on the 

 subject agree in some points, but differ in quite as many. 

 The liquid surface disappears and reappears at different 

 volumes, and the temperature is sometimes found the same 

 always, sometimes depending on the volume. Which 

 volume and temperature is one to choose for the critical 

 values ? And even if we knew which, how is all this to be 

 reconciled with Andrews' conceptions.'^ And this is not all. 

 A number of sometimes ingenious experiments, in U tubes, 

 O tubes and otherwise, have been devised which seem to 

 prove that even above the temperature of disappearance of 

 the liquid surface there is a difference between denser and 

 lighter substance. The liquid, though not separated from 

 the vapour, would seem still to exist and only gradually to 

 become mixed with the vapour. The point where the 

 meniscus disappears would not be the true critical point, 

 but the latter would be found at a higher temperature. 



There is no doubt that observations near the critical 

 point are of the utmost difficulty, the chief reason being the 

 enormous compressibility and expansibility of the substances 

 in that region. It has also been proved that near the 

 critical point exceedingly small admixtures have an enor- 

 mous influence on the density. In an experiment due to 

 Galitzine two quantities of ether, separated by a column of 

 mercury, are enclosed in a U tube. The tube is heated 

 above the critical temperature, and shows differences of 

 density in the two limbs. According to Galitzine, these are 

 due to a greater number of liquid molecules on the one side 

 of the U tube. The writer of this article repeated this experi- 



