( 279 ) 



— 0,685 -1- 3,8G r= - f / ) =3,18. 

 p \d.vJrT 



The fact that l\.v, AT and A/> cannot be considered as differentials 



will undoubtedly contribute to the circumstance that this quantity 



shows such different values if calculated from Keesom's observations. 



1 /dp\ 

 But though the calculated \'alues for — t~ are not the same, 



it appears sufficiently that the value of this quantity lies in the 



neighbourhood of 3, and probably above it. That the equation of 



state gives a so much lower value if we put b constant, must be 



attributed to the fact that the influence of this erroneously introduced 



simplification is great here, whereas this simplification caused hardly 



Td.v„ ^, , 1 /dp\ 



any error in the calculation of ,„, . The value of — ( ^ we 



dl p \0a;J,.2' 



found equal to : 



a i 1 da 8 v^ 1 db) 



v> j a dx 27{v—by b dx\ 



a 

 With Vy.=^^b we find the value 3 for , while the second 



a 

 dl — 



b \ dlb ^ « , 



factor becomes equal to \- ~^——- But it is sufiiciently known 



dw 3 div 



that the critical volume is much smaller than 3 b, and that the 



variability of b accounts for it. The same cause to which it is due 



T /dp \ 

 that at the critical volume — h^T;^ is found equal to 1 + 6, instead 



p\oIJ, 



a 

 of 1+3, causes us to find equal to 6 instead of to 3. Let us 



briefly prove this. 



/dp\_MRT _ a 



^\df)~ v-b ~^ '^V 



p\pTj pv^ 



In the critical circumstances the value of the first member is 

 about 7 or = 6. If we use this value, we find for 



1®- 



double the previous value, i. e. 3,5. The second factor of — ( ^ ) 



