( T95 ) 



that of n ri'ilical point can l)c dcrixptl ' ). it is Iciiown that orpiation 

 (2) iu<lical(>s holli, tlKHiuIi not nniiicricallv accui-alc; one of iIkmhiiuc- 

 roiis casos. wiicic the ('([iialion (•!' \ \n dkij Wam.s is a safe u-uidc 

 ÏOV tlic (pialilali\<' coiirso of llic pliciiomciia. Ilioiiuii il is imal)lo to 

 represent tlieiii (pianlilali\('l\ . i\oirrr,\\ K(; (Icimxcs. tlicrcroi-o, IVoni the 

 ecpiatioii (I). (2)') hv |iiitlinu as il was dt'chiccd Ity van di-.k Waals 



and himself /' = loi' llic \ahi(" ha- th(> innnher ot coliisions 



/• — /; 



instead of /'= , which vahie was nsed l\v Tait an<l 1>ohkntz. 



V 



This disenssion has not \r<\ lo a pei-fect agreement, any more than 

 a Uiter discussion carried on between Boltzmann ^) and van dkk 

 Waals '') about tlie corrections, which ai-e to be applied to the ^•alne 

 of b, which is pnt constant in (1) and (2) and eqnal to the fourfold 

 of the volnme of the molecules. As is known, .l.ürKR') and Boltzmann ") 



found by lirst a})in'oxiniation />,, = /)^ [l -]- ~- j for the A from (1); 



/ 11 b\ 



VAN DEK Waals A,. = A 1 ~ ^; h)r that trom (2): aftei'wards 



* V •^- 'V 



3// 



VAN DKR Waals Ji'. ') has found for the latter hr=/).l I 



in a diiferent way, so that his result, as far as the terms of the order 



and — are concerned, agrees with that of .1.\(;f,h and Bot.tzmann. In 



his |)ublications \vhich have appeared since "), his father has |»ro- 

 nounced himself "inclined to acknowledge "/^ as the correct value," 

 l)ut it is ]U)t doubtful for an attentive reader, that this "inclination" 

 leaxes ample room for (h)nbt, both with regard to the \alue of the 

 coetHicient ^/^, and to the following coeflicient j^, wliich was gi\en 

 on one side as 0.095^, on the other side as O.O.'JHO. 



1) Evidenlly Tait lia.s not seen this, but ho thinks tlint tlie peculiarity of I'oi'ni 

 (2) exists in this, tiial il is a cubic witli respect to r; evidenlly on acconnt of 

 the part which the three sections of the isollierm wiih a line parallel lo the 

 /■-axis, play in the theory of van per Waals. But he overlooked, tliat every valid 

 equation of state will have to represent these three volumes. 



~) See also van deu Waals: (lonliiniitiil ls*.»'.l. p. CtO. 



3) These Proc. I, p. 398. 



*) These Proc. I, p. 408. 



") Wien. Sitzungsber. 105, p. lo. 



*') Gastheorie, p. 1Ö2. 



7) These Pioc. T), p. IS7. 



8) These Proc. Ü, p. 135. 



