48 . THE ROYAL SOCIETY OF CANADA 



to the values of their critical coefficients. In the case of heHum, 

 whose molecules approximate the ideal structure imagined by van der 

 Waals, the criflcal coefficient is 2•68^ a very close agreement with the 

 theoretical value. 



While, strictly speaking, it is not proper to apply the method of 

 corresponding states to mixtures, the application may be justified in 

 the case of air, since its components are not very unlike. We should 

 avoid the critical region of any mixture, however. 



"Previous work^ on the critical phenomena of air has shown that 

 the temperature and pressure of the plait-point are -140-73° and 37 «25 

 atmospheres respectively, and that this point is situated very nearly 

 on the straight line joining the critical points of pure oxygen and 

 nitrogen. Also, the density «36 at the plait-point is almost what the 

 centre-of-mass law would give. Further, the critical point of contact, 

 or highest temperature and corresponding pressure at which liquid 

 is possible is very near the plait-point, so that we may take the data 

 at the plait-point for calculation of the critical coefficient. The 

 value thus obtained is 3-46. This value lies between 3-42 and 3*59, 

 the values for ethylene and carbon dioxide respectively. This sug- 

 gests the comparison of the isotherms of air with these two gases. 



It was discovered very early in the study of the isotherm that the 

 curves for various substances were of the same general form. Amagat^ 

 showed that if scales proportioned according to the critical pressures 

 and volumes were used, the isotherms of different substances fell into 

 the same family. When the proper scale has been determined, the 

 critical data of a gas may be estimated from those of any other sub- 

 stances. 



The method used by Raveau* is a distinct improvement for the 

 purpose of determining critical constants. Instead of plotting the 

 isotherms in terms of P and v, he used log P and log v. Thus the 

 isotherms of one substance may be laid on those of the other and 

 shifted until they come into agreement, the co-ordinate axes being 

 kept parallel. Just as Amagat found the Pv, P diagram more 

 convenient than the P, v diagram for studying the departure of 

 the isotherm from the simple gas law, so the log Pv, log P diagram 

 is more convenient than the log P, log v diagram, being more 



Pv 



compact. The log —, log P diagram is still more compact and has 



^ Onnes, Communication from the Physical Laboratory of Leiden, 119 (b), 1911. 

 ■^Communications, Leiden, 150b, 1917. 



Transactions of the Royal Society of Canada (III), XI, 1917. 

 ' Amagat, Statique Expérimentale des Fluides, Rapport présentés au Congrès 

 International de Physique, Tome 1, 551, 1900. 



* Raveau, Compt-Rend., 123, 109, 1896. Jour, de Physique (3), 6, 432, 1897. 



