and its relation to Latent Heat. 3 



extent of liquid surface. The work performed from one to an- 

 other of these contours was taken as the mean volume of the 

 liquid descending through the distance of the centres of gravity, 

 and the corresponding surface denuded was the difference of sur- 

 faces at those times. 



The contours of tube A corresponding to the highest and low- 

 est position of the centre of gravity thus integrated and computed, 

 gave 132'3 as the value of the quotient of the surface denuded 



by work performed. Expressed by symbols, this quotient Q = tq-> 



in which a is the square root of difference of surfaces^ b-tlas diih^ 

 root of mean volume of liquid, and c the vertical distance between 

 the centres of gravity, all expressed with reference to an inch as 

 the unit. 



Two adjacent contours of tube B, similarly computed, gave 

 Q=126. Two contours of B further removed, similarly com- 

 puted, gave Q=134. 



The atmospheric temperature at which these experiments were 

 made was 86° (the locality being within the tropics), but the 

 temperature of the surface of the drop was somewhat uncertain ; 

 first, the lamp being within a few inches, tended to heat it ; and 

 secondly, the atmosphere being very dry, the evaporation from 

 its surface, sensibly rapid, tended to cool it. 



Thus far it seemed proved that the tensile force at the surface 

 of water is uniform, and that the contour of the suspended volume 

 is determined through its complete cycle of elegant curvatures 

 by the quotient of the differential of surface (upper and lower 

 inclusive), by differential of work being a constant quantity. 



§ 3. The maximum volume suspended at the extreme develop- 

 ment of the drop just before it separates, was found in the case 



rj.AQO 



of tube C of the dimensions _. ■.^„ , by the same method of obser- 

 0-129 •' 



ration and graphical integration as employed with tubes A and B. 

 The quotient of the sum of the outer and inner rims of suspen- 

 sion by the volume suspended was 129. 



The same tube C held vertical was brought down until its 

 lower rim just touched the water in a cup, and the volume sus- 

 pended by the inner surface of the tube computed : the quotient 

 of the length of the inner circumference by this volume was 127. 

 The temperature was about 86°, as before. 



§ 4. Su])pose a strip of paper, one inch wide, immersed in 

 water and its whole surface wetted. In pulling it out we deve- 

 lopc or denude an aqueous surface of two square inches for <'ach 

 inch of vertical height; and the work recjuired to effect this 

 would be ascf;rtaiu(;d if we could woigli the water adhering to 

 the lower edge of the paper at the instant of separation ; for this 



B2 



