314 Mr. W. Sutherland on the 



valid, still a ocpa. The seeming discordance between these 

 two values for a corresponds to the old difficulty in the early 

 days of capillary theory, when Laplace's simplification of 

 imagining the surface-density to be the same as the body- 

 density was adopted, and the surface-tension was found expe- 

 rimentally to diminish much more rapidly with temperature 

 than his theory indicated. But if we take account of the 

 possibility of a markedly different density in the surface from 

 that in the body, and recognize the corresponding difference 

 of stress in the two regions, we shall be prepared to imagine 

 that, approximately, p^ ozp^, which means that the rate of 

 variation of the surface-density with temperature is more 

 rapid than that of body-density. And this is a necessary 

 corollary of our concession of a marked difference of density 

 and stress in the two regions ; for Van der Waals has accus- 

 tomed us to the idea of a difference of stress, of the order of 

 magnitude of 1000 atmospheres ; and the recent researches of 

 Amagat, on compressibility at higher pressures and its varia- 

 tion with temperature, have shown that the coefficient of 

 expansion of aather under one atmosphere between 0° and 50° 

 is '0017, while under 3000 atmospheres it is '00056, so that 

 an increase in stress of 3000 atmospheres diminishes the 

 coefficient to one third of its ordinary value. Thus then we 

 see that, in a general way, the old difficulty is really a very 

 direct confirmation of the great difference of stress and density 

 in the surface-film and body of a liquid ; and moreover ex- 

 periment shows that heating the film alone produces a much 

 more pronounced effect than heating the lower liquid only in 

 a capillary tube. 



We can now proceed to determine the values of the para- 

 meters kA for a large number of compounds by applying the 

 equation a = kAp^m's to Schiff's results. 



The following table is devoted to the consideration in the 

 first case of the values of kA for the isomeric forms of the 

 compound ethers of the general formula Gnii2n02- For the 

 sake of brevity the basic radicals methyl, ethyl, &c. are de- 

 noted by I, II, &c., and the basic radicals in ascending order by 

 1, 2, 3, &c. ; so that 1 1 stands for methyl formiate, II 3 for 

 ethyl propionate, and so on. Schiff's values of a are given 

 in terms of grammes weight per linear metre, so that ^A in 

 the following tables is given in corresponding units. The 

 acids CnHsnOg are omitted from the table because Eotvos 

 finds that they do not obey his law, and their anomalous 

 vapour-densities are well known. 



