264 SCIENCE PROGRESS. 



phenomenon at 15 '6' C. (/p), retrograde condensation with 

 decreasing" maximum quantity of liquid from 15 '6° C. to 

 19° C, and no condensation above 19° C, the critical 

 temperature 4. 



The same phenomena have been observed with a 

 number of other mixtures, such as mixtures of carbonic 

 acid and methyl chloride. The Interval between /p and 4 is 

 however often much smaller than it was in this case. In 

 mixtures of carbonic acid with acetylene and nitrous oxide 

 with carbonic acid and with ethane the distance 4 ~ ^p ^'^'3-s 

 so small that no retrograde condensation could be detected 

 at all. These mixtures behave almost like single sub- 

 stances. This result may also partly be due to the action 

 of gravitation which disturbs the real phenomena for mix- 

 tures in the way just described. 



We are now in a position to complete the p v diagram 

 for mixtures and consider in what respects it differs from 

 the same diaofram for single substances. The isothermal 

 ^=io°C. was considered before. The curve shows two 

 breaks at b and e : in experiments on mixtures of carbonic 

 acid and sulphurous acid Blumcke found the isothermals con- 

 tinuous even below the critical region; this is due to retarda- 

 tion which tends to blur the discontinuities at b and e. 

 On stirring the mixture the discontinuities reappear in the 

 diagram. Through the different points b and e belonging 

 to the isothermals at different temperatures we may as be- 

 fore draw a curve, the border or saturation curve. The 

 isothermal for the critical temperature 4 is then found to 

 touch the border curve, but not as for sinole sustances at 

 the top M of the border curve but at a point C to the right 

 of M. The critical pressure is not the maximum pressure 

 on the border curve. The isothermal for t^ the temperature 

 at which the mixture shows the critical phenomenon at P 

 has a shape similar to the shape of the isothermal /= loX. 

 Sometimes P is found on the left of M, sometimes on the 

 right, and in exceptional cases it coincides with M. The 

 isothermals which meet the border curve, must be supposed 

 as for single substances to exist also inside the border curve 

 and the same experimental facts exist to support this hypo- 



