26o SCIENCE PROGRESS. 



compressing the mixture at a temperature below ig°C., say 

 at i8°C., an entirely new phenomenon is observed. The 

 liquid appears as at io°C. at a certain volume (point d of 

 the isothermal) and the pressure goes on rising after that ; 

 the liquid increases at first but its surface becomes less 

 and less sharply defined and long before the whole of the 

 mixture is liquefied the surface becomes a mist and dis- 

 appears altogether in very much the same manner as the 

 liquid surface of a single substance disappears at its critical 

 point. On lowering the pressure the mist and the liquid 

 surface reappear at about the same level where the dis- 

 appearance took place before, and further lowering of the 

 pressure gives the phenomena of compression in the oppo- 

 site order. Similar phenomena are observed at lower 

 temperatures, but the lower the temperature the larger the 

 quantity of the liquid phase is when the surface disappears 

 or reappears. Below I5°C. the whole of the mixture is 

 liquefied, as described for the temperature of io°C. If we 

 call the disappearance or reappearance of the liquid surface 

 in a mist the "critical phenomenon" we may express the 

 experimental results by saying that the mixture shows the 

 critical phenomenon not at one temperature only, but over 

 a range of temperatures of about 4°C. (i5°-i9°C.). By 

 saying that 19°C. is the critical temperature for our mixture 

 we evidently do not exhaust the phenomena. Above iq°C. 

 no condensation takes place, but the critical phenomenon 

 is not confined to that temperature but occurs at lower 

 temperatures also. In fact as will appear presently the 

 critical phenomenon really belongs to a temperature lower 

 than i9°C. (about i5"6°C.) and it is in consequence of retar- 

 dation and gravitation that it also occurs at temperatures 

 up to the critical temperature. 



Phenomena like those sketched were also obtained by 

 Andrews with mixtures of carbonic acid and nitrogen, by 

 Van der Waals with carbonic acid and hydrochloric acid, 

 and others. Andrews' experiments are the most complete. 

 He noticed great irregularities in the behaviour of his 

 mixtures and thoroughly investigated their cause. He 

 showed how they are explained by imperfect mixing of the 



