10 Mr. J. J. Waterston on Capillarity 



observations with parallel plates, it would be desirable to have 

 a set of observations made with water between plates at atmo- 

 spheric temperatures between summer and winter, in the manner 

 adopted by l\I. AVolf. 



Tliat heat has less cflfcet on the plate-column than on the 

 tube-column, is the impression I have from many oliscrvations. 

 When we consider tliat the effect of adding a very small propor- 

 tion of alcohol to water is several times greater in tubes than 

 between plates, it appears as if any cause that increased the lim- 

 pidity of the water had more effect on the capillarity of tubes 

 than of plates. 



§ 12. In § 4 it was shown that the weight of 38 grains de- 

 scending through one inch represents the work required to denude 

 20 square inches of aqueous surface at the temperature of 8G°F., 

 if the value of Q is 132-9. If it is 132, as assumed from the 

 observations in § 2, the weight must be taken at 38'2G2 grains 

 descending through one inch ; and to denude one square inch, 

 it requires 1-9131 grain descending through one inch vertical. 

 This being the weight of -007577 cubic inch of distilled water at 

 the temperature of 86° F., this volume of water raised through 

 one inch is equivalent to the work of denuding a superficial stra- 

 tum of one square inch, of overcoming the integral cohesion force 

 on one side — the outer side — of a superficial stratum of molecules; 

 being one-sixth of the cohesion integral of all the molecules in 

 the stratum. 



A molecule at the surface of a liquid differs from a molecule 

 below the surface, in having that part of its cohesion force which 

 is directed outwards, unmated — disengaged ; and the force that 

 operates in disengaging it, is the weight of liquid held in capil- 

 lary suspension. On applying heat to water to convert it into 

 vapour, we overcome the cohesion force of all the molecules, and 

 the quantity of work which this is equivalent to, we can readily 

 compute from the data afforded by M. Ecgnault and Mr. Joule, 

 assuming the latent heat of steam to be the cohesion integral of 

 all the molecules while in the liquid state. Thus, having the 

 cohesion integral of one stratum of a cubic inch, and the cohesion 

 integral of all the strata of a cubic inch, we obtain the number 

 of strata in an inch, and hence the absolute volume of an aqueous 

 molecule. 



With similar data for other liquids we may compute the abso- 

 lute volume of their molecules. 



Now the quotient of atomic weight by specific gravity of one 

 liquid compared with the similar quotient of another, gives us the 

 ratio of the absolute atomic voluni' . . 



We have thus two distinct modes of computing this ratio, — 



