96 Dr. R. D. Kleeraan : Determinations of the Law of 



substances decrease. On the whole, it appears that the 

 method is of little value to determine the law of molecular 

 attraction. 



The arbitrary function in the law of attraction given by 

 the writer consists possibly of the sum of a number of 

 positive and negative terms. In that case the force between 

 two molecules will change in sign a number of times as they 

 are brought nearer to one another. This could be experi- 

 mentally investigated by allowing a monatomic gas to 

 expand without doing work, and measuring the cooling or 

 heating effect produced (Joule-Thomson effect). If a heat- 

 ing effect is produced there is repulsion for distances of 

 separation of the molecules lying between their distances of 

 separation in the two stages ; if a cooling effect is produced 

 the force is one of attraction. It seems best to use a mon- 

 atomic gas, as we cannot be sure that the frequency of 

 collision of a complex molecule — w<hich changes in the above 

 process — alters the configuration of its atoms and thus 

 causes the effect observed. It is interesting to observe that 

 the distances betw r een two molecules of the same kind for 

 which the force changes sign depends, according to the form 

 of the arbitrary function in the law of attraction, on the 

 nature of the molecule. Tf tw T o different pairs of molecules 

 are at corresponding temperatures and the molecules of one 

 pair are separated by a distance z\ and those of the other 

 pair by a distance z", the forces will have the same sign 

 when 



that is, when the distances are to one another as the 

 distances of separation of the molecules at the critical state, 

 for this makes the values of the arbitrary functions assume 

 the same sign in the two cases. It follows, therefore, that 

 the distances between tw r o molecules of a different kind for 

 which the force changes sign are not the same as those 

 when the molecules are of the same kind. The Joule- 

 Thomson effect of a mixture of gases should therefore not 

 be an additive property of the constituent gases, and this 

 has been found to be the case *. It will also be easily 

 seen that the sign of the Joule-Thomson effect should 

 depend on the density of the gas before and after expansion, 

 and keeping the conditions of expansion constant it should 

 depend on the temperature. When a mass of gas expands 



* Preston's ' Theory of Heat/ p. 811, second edition. 



