( 1fi2 ) 



is jïreater than the isopentane volume, and this is in agreement \Yith 

 the \'aliies which we found for K^ and ^/.m viz. 



/iT/argon) =3.28 while 



K^ (isopentane) = 3.73 ; 



/^?M (argon) = 0.9027 while 



i^M (isopentane) ^ 0.8923 ^). 



The most obvious explanation of these larger values for argon is to 



ascribe tliem to the absence in argon of the usual coni|)ressibility of the 



molecules of normal substances.^) Indeed, taking foi- the moment the van 



Ro 

 DEK Waals isotherms ).:= — that is to say, a constant (i.e. independent 



8/' 



of the temperature) displacement of loij ^.o means that h is increased 

 or decreased in a definite ratio. Looked at from this point of view 

 the accompanying diagram gixes immediate experimental expression 

 to the fact that the argon molecule is le.ss compressible than the 

 molecule of an ordinary normal substance. From this it is probable 

 that the differences between the values of A', and of a-s] must also 

 be ascribed to the same cause. 



The question still remains if we ai'e here dealing with an actual 



^) It is worth noting the appearance of the vapour branch for argon which, in 

 tliis diagram, ahiiost coincides with that for isopentane. If we imagine a reduced 

 GiBBS surface on li.e one hand for argon, on the oilier for isopentane, then for 

 argon the liquid crest is displaced towards the side of larger volumes, and in the 

 vapour region the surfaces fake a dilTcrent form on account of the deviations of 

 the isotherms. The two causes, therefore, in conjunclion v/ith a raising of the 

 liquid crest must so operate that a coincidence of the vapour branches of the 

 boundary curves is obtained. 



■-) As will be shown in an article by H. Kameuli.ngh O.nnes and W. H. Keesom 

 on the equation of state and its graphical treatment which will shortly appear in 

 the Encijdojiaedie der mathematisrhnn Wissenschaften, dilïercnces between 

 the surfaces of state for various substances such as have here been found to 

 exist between argon and isopentane may also be brought about by a difference 

 in tlic distribution round the molecule of the mean attraction potential, for 

 example, by a diflerence in the ratio between tlie hairing distances (i. e. distances 

 over which tlie potential is diminished by half) to the radius of the molecule 

 supposed spherical with which, according to the latest considerations put forward 

 by VAN DER Waals, should be connected a varying degree of a|iparent association. 

 On the other hand sucli a difference in distribution of the attraction potential could 

 depend on a difference in the molecular shape ; it might, for instance, be that 

 with spheroidal molecules either prolate or oblate the possibility of clcser packing 

 may give rise to an overlapping of more concentrated parts of the corresponding 

 fields of force, which would conduce to the development of a greater attraction 

 virial (and consequently also to a higher value of the critical temperature.) 



CI', also Proc. March 1911, Comm. N '. 119. 



