SURFACES OF DISCONTINUITY 597 



would be identical with that in the interior, but with lapse of time 

 the molecules affording the greater surface tension would tend 

 to leave the surface and a fall in surface tension would be 

 observed. In his book (page 152) Adam criticizes the experi- 

 mental evidence which has been brought forward to substantiate 

 the hypothesis on which this argument is built, pointing out 

 that the purely mechanical effects of the appliances employed 

 could easily account for the initial elevation of water in a 

 capillary tube apart from the effect of the postulated increase 

 of surface tension at the beginning. He states that until 

 apparatus capable of dealing with liquid surfaces not older than 

 0.005 sec. has been devised, the question cannot be regarded as 

 settled. Undoubtedly contamination produces change in 

 surface tension. 



XIII. The Influence of Gravity 



S7. The Variation of p, a, m, p.2, • • • with Depth in a Liquid. 

 An Apparent Inconsistency in Gibbs' Argument. The 



Argument Justified 



Before proceeding to consider the question of stability of 

 equilibrium it will be well to dispose of the subject of equilibrium 

 in itself and to proceed at once to deal with the subject matter 

 treated on pages 276-287 of Gibbs' treatise which is the natural 

 continuation of the considerations raised earlier on pages 144- 

 147. The conditions obtained there still hold, with the addi- 

 tional important equations [614] and [615]. There is a certain 

 economy in the notation at the outset of this subsection. In 

 [599], for instance, djDe^ stands really for the sum of a number 

 of terms such as 



8fDe'y + SfDe"^ + SfDe'"^ + . . . 



each referring to one homogeneous mass, while 8j^gzDm^ is a 

 double sum such as 



dfgzi' Dm,' y + bfgzi' Dm-l y ... 



+ 8fgz,"Dm/'y + hfgz^'Dm^'y . . . 



+ etc. 



