Theory of Surface Forces. 349 



the departure from the law of Pascal in a curved capillary 

 layer* : 



H = - 



^{V*-* £?♦£[£?;•*»* o» 



where \ and a are constants, while denotes the thermic 

 pressure in the considered point of the capillary layer. For 

 a plane capillary layer pi and p 2 are respectively the reciprocal 

 values of the abscissae of the points H and K in fig. 1, and 

 therefore the ordinary densities of the liquid- and vapour- 

 phases. In our case, however, where the capillary layer is 

 curved, the limits of the integral in the formula (19) are the 

 densities of the points A 8 and C 8 instead of those of the points 

 H and K. Now the thermic pressure, which is considered in 

 the capillary layer as a pure function of the density (Stefan, 

 Fuchs, Rayleigh, van der Waals), increases with the density. 

 The contribution to the terms of the considered integral is 

 therefore the most important in the neighbourhood of the 

 points H and A 8 , while the contribution corresponding to the 

 points between C 8 and K is much less important. 



The difference of the abscissae of the points A 8 and H being 

 very small, we see thus that the order of greatness of H given 

 by the expression (19) is the same for a curved capillary 

 layer as for a plane capillary layer. For the considered 

 temperature T = 0"844T t or 121°-5 Celsius the capillary 

 constant of Laplace is for ether : 5*17 dynes per cm. The 

 equation (18) gives thus : 



R = tttk^ x 10 -6 = about 5 x 10~ 7 cm. = 5u/x. 



R being a mean value between the two spheres which limit 

 the spherical capillary layer, we find for the radius of the 

 sphere which limits the drops a value between 5 and about 

 8 fifi, while my formula (14) for the thickness of the plane 

 capillary layer of ether gives at 121°' 5 Celsius, h = about 10 /a/jl. 

 We find thus actually for the radius of a liquid drop of ether 

 which has its minimal size a value of the same order of great- 

 ness as for the thickness of a plane capillary layer at the 

 same temperature. 



Secondly, we make the corresponding calculation for a 

 liquid drop of ether at the temperature of 0° Celsius or 



T 

 a = ™- = 0-585. 



* Phil. Mag-. Dec. 1906, p. 5G6. The demonstration of the formula 

 for curved capillary layer is quite the same as that at the locus citatus. 



Phil Mag. S. 6. Vol. 17. No. 99. March 1909. 2 B 



