28 THE TRUE VALUE OF a OF VAN DEE WAALS' 



method involves the assumption that the angle of contact is zero 

 and, while this is undoubtedly very nearly true at low tempera- 

 tures, the angle has a finite value at higher temperatures. The 

 correction, if made, would have the result of increasing the surface 

 tension. The surface tension of all the liquified gases given in the 

 table, except that of SO.-,, were determined by the capillary rise 

 method and are hence to be regarded as minimum vaines. In S0 2 

 the tension was measured by the ripple method. In most of the 

 other cases cited the tension was measured by the drop weight 

 method, and I have used the results of Morgan l ) and his col- 

 leagues , but I have corrected his results by the method of Loiin- 

 stein 2 ) and Harkins and Humphery 3 ). In interpreting the values 

 of a found in this way it must also be borne in mind that if 

 there is any association in the liquid the effect is to reduce the 

 value of (', for the reason that the number of molecules does not 

 then remain a constant but is reduced by the association. This 

 again will in some cases make the value of a lower than it 

 really is. In general then the values of a determined by the 

 surface tension are apt to be lower, rather than higher, than the 

 true values. The fact that C is constant over a wide range of tem- 

 perature shows that van der Waals' constant a is also constant 

 as he has maintained, but which has been doubted by others. 



I may mention, also, in this connection that if the density at 

 absolute zero was really equal to Sd c , where d c is the critical 

 density, the constant C would have the value ascribed to it in my 

 paper on the Internal Pressure of Liquids of C= 7i',.7//3/ J ( JV 1/3 , 

 where N is the number of molecules in a gram mol and K c is 

 a\V^. If K ( .\V t . were constant in all non-associating substances and 

 had the value I have assigned to it of 0.5, then C would be the 

 same in all normal substances and it would have the value of 2.14, 

 which is very close to the value found for it in many substances. 

 However as d is not always equal to Sd c , this value of C is not 

 always the same, but is lower in the simple' gases and higher in 

 the more complex substances, ranging from 1.60 to 2.23 or perhaps 

 still higher. 



The values of a calculated from the surface tension are given 

 in column 6 of table 2. Their agreement and disagreement from the 

 other methods of calculation will be readily perceived from the table. 



i) Morgan et al: Journal of the Amer. Ghem. Soc, 33, .34, 35, (1911), (1912), (1913). 



2) Lohnstein: Zeit. phys. chem., 64, 686 (1903); 81, 410 (1915). 



3 ) Harkins and Humphery: Jour. Amer. ('hem. Soc., 38, 228 (1916). 



