Surface Forces in Fluids. 355 



of the liquid, and the portion G H K, which we know to 

 represent conditions of the vapour, are not really discon- 

 tinuous, but are joined by a portion C D E F G, of which the 

 part between D and F must be unstable — is very similar to 

 our own argument from the two curves. 



The advance that we have been able to make on Professor 

 Thomson's suggestions is, that we have shown that it is a 

 necessity of the equilibrium of the liquid that near the sur- 

 face it shall be less dense than is indicated by the point C, i. e. 

 that the curve does pass below this point. Further, that, in 

 the case of liquids whose vapour-pressure at the given tempe- 

 rature is not enormously great, there is near the surface a 

 region of true tension — that is, that the curve passes below the 

 horizontal axis of coordinates in Maxwell's diagram; while by 

 the argument from the breaking-strain we see that, after a 

 certain volume has been reached, the corresponding tension 

 diminishes, i. e. we have experimental evidence of the existence 

 of the curve down to and even beyond the turning-point D. 

 Again, by showing that it is a necessity of equilibrium that 

 the vapour should be condensed at the surface of the liquid, 

 we have been able to prove that the vapour really exists there 

 in the condition indicated by the curve between G and F. 



Thus Professor Thomson's suggestion (see Maxwell's 

 i Heat,' p. 127) that such a state of things as is indicated by 

 his hypothetical part of the curve " may exist in some part 

 of the thin superficial stratum of transition from a liquid 

 into its own gas, in which the phenomena of capillarity take 

 place," is shown, by our method of considering the question, 

 to have been a correct anticipation. 



We may now observe, with respect to the hypothesis which 

 we have used throughout, viz. that the repulsive force is only 

 exerted between contiguous layers of molecules and not be- 

 tween more distant layers, that this conception of the force 

 is the only one by which we can represent statically the phe- 

 nomena of impact by which we believe the repulsive action 

 to be really due. The choice of this hypothesis is, in fact, 

 one of those to which we are guided by our knowledge of 

 the dynamical phenomena ; and this fact itself, and the com- 

 pleteness with which the hypothesis serves to express the 

 phenomena, points to the correctness of the choice, which 

 could be further justified by showing that any more general 

 hypothesis would not express the phenomena without the 

 introduction of complicated restrictions. 



15. The next step which it is necessary to take is to show 

 that the investigation we have given would be in no way 

 affected if the surface of the liquid were curved instead of 



