CLARK. SURFACE TENSION BETWEEN LIQUIDS AND VAPORS. 363 



g H(d 



where a is the surface tension, H the height of the meniscus, d and d' 

 the densities of the liquid and vapor respectively, w the contact angle, 

 and r u r 2 , and r s , the inner and outer radii of the capillary tube and the 

 inner radius of the containing tube respectively. The angle of contact 

 was found by measuring the depth of the meniscus and the radius of 

 the tube. Then considering the meniscus a portion of a sphere, 



w = - — 2 tan ' — , 

 2 r x 



It being the depth of the meniscus. This method of determining a> is 

 not satisfactory, for it assumes a spherical meniscus, which shape is of 

 course realized less and less as we near the wall of the tube. It also as- 

 sumes a perfectly circular cross-section. A better method has been pro- 

 jected, but not yet tested. 



The capillary tube method of measurement of surface tension leaves 

 much to be desired when applied under such severe conditions as in the 

 present work. 



The tube used in the liquid experiments was not of the complicated 

 form used in the vapor tests, but both liquids were placed in the same 

 tube. The measurements for the determination of a were made exactly 

 as just described. 



These tubes were placed in an electric heater controlled by an auto- 

 matic thermostat, and the constant observed for different temperatures. 

 The lower part of the experimental tube was connected with a manom- 

 eter. In this way the capillary constant and pressure could be deter- 



* The sjstem being in equilibrium, the pressures inside and outside the capil- 

 lary tube at the level of the meniscus outside are equal. 

 So 



2 it ri T cos w-Ttr^r/d l _ 2 7T (r 3 + r 2 ) T cos to - ir (r 3 2 - r<?) g II d. 2 

 irr-f M?- 3 2 — r 2 2 ) 



or 2 T cos w(~ — ) = q II (rf, - d 2 ) 



VI r 3 ~ r l' 



„ q H (d x - d„) 

 whence T = = \~ ^ 



2 cos o I ) 



V'l r 3 -r 2 J 



