Thermal Properties of Carbonic Acid and Nitrogen. 439 



Soon after the publication of the posthumous paper, Herr 

 Margules* calculated some thermodynamical magnitudes 

 based on the data supplied. Herr Galitzene f developed the 

 incomplete result of comparison made by Andrews himself 

 between the experimental data and Dalton's law of gaseous 

 mixtures. 



Characteristic Equation. 



We are not at present in possession of a rational form of 

 characteristic equation applicable to all substances. Some 

 of those which have been proposed as such hold good only for 

 one class of substances, nor do they continue to hold strictly 

 through the whole range of physical changes within the reach 

 of experiment even in the case of the very substances for 

 which they were particularly given. They are therefore to be 

 looked upon as empirical and tentative, and we have to make 

 use of them as such for the time being, or to regard them as 

 interpolation formulae whose mathematical forms are much like 

 those of a rational one. In choosing the Clausius* form of 

 characteristic equation, I was guided by no other consideration 

 than the above. It is 



_ _rt_ e 



in which © is a certain undetermined function of T, the 

 simplest form being 



~ K (a constant) 

 <H> = fp — . 



In the first place I have to determine the four constants 

 E, K, a, and /3 for the mixture under consideration. 



As Andrews used a hydrogen-manometer, and hydrogen 

 presents a decided deviation from the gaseous laws, cor- 

 rections for the given manometric values of pressure are to be 

 applied. It is certain, although not explicitly mentioned, 

 that Andrews calculated pressures by the equation^ 



Vo(l+«Q g 

 p ~ Y 1 ~*"760' 



in which V is the volume of the mixture at 0° and under the 

 normal pressure, Vj that at the temperature t and under the 



* Wiener Sitzungsbericht, Bd. xcvii. (1888). 

 t Wied. Ann. Bd. xci. (1890). 

 \ ' The Scientific Papers,' p. 422. 



