562 Mr. S. A. Shorter on Surface Separation from 



Theory of the Method. 



Suppose that the solution is contained in a cylindrical 

 vessel of radius R, and that a circular brass disk of radius r 

 is suspended by means of a "fine wire, so that the axes of the 

 vessel, disk, and wire are coincident. If the solution fills the 

 vessel to such a height that the disk is partially immersed, 

 then, neglecting the capillary curvature at the edges, the 

 surface pellicle will be in the form of a thin annular lamina. 

 Let z be the thickness of this lamina and n the rigidity of 

 its material. It is easily shown that a rotation of one of 

 the circular boundaries with respect to the other about the 

 common axis, will give rise to a couple of magnitude 



£7rnzWr 2 

 RP-r 2 ' 



Hence, if I is the moment of inertia of the suspended 

 system and T the period of the oscillations set up by an 

 initial rotation of the vessel, we have, neglecting the elastic 

 forces of the wire and the inertia of the liquid set in 

 motion, 



nz — 



RV 2 T : 



In the absence of any knowledge of the value of z we may 

 write 



where fi may be termed the surface-elasticity. We may also 

 write 



A 

 where A may be termed the " apparatus constant." 



Apparatus. 



Two forms of apparatus were used in the following ex- 

 periments. The first was identical with that used in my 

 previous experiments on old surfaces of saponin solutions. 

 In the second form the suspended system was hung from 

 supports attached to the wooden base which carried the 

 turntable on which the vessel rested — the whole being placed 

 in a wooden box. The bottom of the box was perforated so 



