and Surface Energy, 241 



p. 276 ; also Stefan, Wied. Ann, xxix. 655). Seeing that 

 the internal latent heat of a liquid is presumably a measure 

 -of the work done against the internal pressure, and that the 

 work done by molecules in getting into the surface of a 

 liquid (i. e. halfway out of it) is measured by the potential 

 energy they acquire there as surface energy, it is argued 

 that one-half the latent heat must be equal to the molecular 

 surface energy, which is proportional to pY* *. 



Apparently no attempt has up to the present been made 

 to give the above general relationship precision or to apply 

 it to the data. In the present communication a definite 

 connexion between the two magnitudes (internal latent heat 

 and surface energy) is put forward and shown to be in very 

 fair agreement with the available data. 



As has been pointed out by Matthews (Jour. Phys. Chem. 

 1916, xx. 555), a molecule in passing into the surface layer 

 of a liquid do8& work only against the component of the 

 internal pressure directed perpendicularly to the surface 

 towards the interior, the lateral attractions on the one side 

 of the molecules being balanced by identical attractions on 

 the other. It would thus appear that the work done by a 

 molecule in getting into the surface layer is not one-half the 

 work it must do in passing altogether from the liquid state, 

 but i x \ or one-sixth, since it is doing work against one 

 only of the three components of the internal pressure. We 

 can therefore conclude that the work done in passing all the 

 molecules in a gramme molecule of a liquid into the surface 

 is equal to one-sixth of the internal latent heat, i. e. to one- 

 sixth of the work to be done in moving all the molecules 

 apart from one another until the liquid is transformed into 

 vapour. 



We have now to devise a method for measuring the 

 potential energy the molecules will have acquired when they 

 are in the surface layer — in other words, we have to deter- 

 mine the true molecular surface energy or energy due to all 

 the molecules in the gramme molecule. We can do this by 

 imagining a gramme molecular volume V of liquid to be 

 spread out in a layer of thickness equal to the molecular 

 diameter on the surface of excess of the liquid. If we 

 assume as a first approximation that the molecules in such a 

 layer are practically contiguous and approximately spherical, 



* Van tTIoff (loc. cit. above) actually equates p\*= \ Jl . 



