224 SCIENCE PROGRESS. 



give ratios which are very nearly the same, but this is not 

 the case for members of Groups II. and III. Now it has 

 to be remarked that these latter liquids are all hydroxy 

 compounds, and we can therefore conclude that if van der 

 Waals' theory of corresponding" states be tested by means 

 of compounds other than hydroxy compounds satisfactory 

 results are obtained. 



Another deduction from theory is that if we divide the 

 density of a substance at the critical point by the theoretical 

 density deduced on the assumption that it behaves as a 

 perfect gas, we ought to get a constant quantity. In the 

 case of Group I. the value of the ratio was almost the same 

 for each liquid, the mean being 376 + '04 ; for Group IV. 

 a similar result held, the mean being 3 "91 + '03. If, there- 

 fore, we confine ourselves to the same group of substances, 

 the above deduction is very closely obeyed, although a 

 liquid in Group IV. gives a ratio which is about 5 per cent, 

 higher than that of a liquid in Group I. The ratios furnished 

 by hydroxy compounds are higher still, the alcohols giving 

 4* 1 9 + "19, and acetic acid 5*00. The exceptional behaviour 

 of these substances is again apparent. 



Now it can readily be shown that on neglecting quan- 

 tities of the second order, the value of the ratio of the 

 critical to the theoretical density, according to van der 

 Waals' theory, should be 2*67. If we take the mean of the 

 values given by members of Groups I. and IV. we get 

 3*84. There is thus a wide discrepancy between the 

 observed and calculated numbers. Some years ago, 

 however, O. E. Meyer pointed out that the factor J 2 ought 

 to be used in connection with several of the quantities 

 employed in van der Waals' theory (5), and Guye (6) was 

 the first to show that 2*67 * ^ 2 or 377 is a number almost 

 identical with the above value deduced from Young's 

 •observations on non-hydroxy compounds. The lesson 

 which seems to flow from this work is that the agreement 

 between theory and practice depends to a marked extent on 

 the nature of the substances, and that the value of the 

 generalisation of van der Waals can only be satisfactorily 

 gauged when this influence has been more fully studied. 



