CHEMISTRY: W. D. HARKINS 
567 
very simple equation, y/v^^^, where V is the molecular volume, should 
give results which lie in somewhat the correct order. This is un- 
doubtedly because, as shown by Harkins, the contribution of the 
kinetic energy of a molecule to the total energy of the surface, is on 
the whole, independent of the nature of the molecule, — at least for 
such substances as have been used in the calculation of cohesion, — and 
is dependent on the temperature alone. Therefore, so long as the 
molecular volume is nearly the same, and the orientation of the mole- 
cules in the surface is not an important factor, at any definite tem- 
perature the latent heat of the surface is nearly independent of the 
nature of the substance, so that when substances are arranged in the 
order of their cohesional surface work or their free surface energy, they 
are also arranged in the general order of their total surface energy. 
A second method of calculating the cohesion from the cohesional 
work or free surface energy, is based on the well-known assumption of 
Stefan,^^ that the average work involved in bringing a molecule to the 
surface of a liquid, has one-half of the energy value involved on the 
average in its complete vaporization. It is obvious from Stefan's 
paper, that his principle does not involve the work but the total surface 
energy, which is supposed to be one-half of the latent heat of vaporiza- 
tion. That this rule is far from true is indicated by the results of ex- 
tensive calculations by Mr. L. E. Roberts and myself, which show that 
the fractional contribution of the surface energy toward the complete 
vaporization increases with the temperature, with a normal range of 
from one- third at lower temperatures to 0.8 or more as the critical 
temperature is approached, though the higher values are uncertain. 
Thus a molecule which at a high corresponding temperature passes from 
the body of the liquid into the surface, goes, in a fractional sense with 
reference to energy, much more nearly into the vapor state than when 
the corresponding temperature is low. 
Negative surface energy. — The phenomenon of negative surface energy 
was first discovered two years ago by Dr. E. C. H. Davies and me, 
but has not been announced previous to this time in print. Not only 
Donnan, but also Tolman and Wolfgang Ostwald, have assumed the 
existence of a negative surface tension or free surface energy. My own 
investigations have convinced me that the discovery of a negative free 
surface energy for a plane, uncharged surface is improbable, though it 
is quite likely that there is such a phenomenon in the case of highly 
curved phase boundaries. What we have to announce here is the 
discovery of a negative total surface energy for a plane, uncharged sur- 
