PHYSICS: I. LANGMUIR 
251 
^Braus, H., Anat. Anz., Jena, 26, 1905, (461). 
6 Harrison, R. G., /. Exp. ZooL, Philadelphia, 4, 1907, (254). 
7 Tornier, G., Arch. Ent.-Mech., Leipzig, 20, 1905. 
sValle, P. deUa, Napoli, Boll. Soc. Nat., 25, (1911-12), 1913. 
^In calculating percentages only those cases which yielded positive results were taken 
into account. 
^0 Bateson, W., Materials for the Study of Variation, London, 1894, (479). 
THE SHAPES OF GROUP MOLECULES FORMING THE SURFACES 
OF LIQUIDS 
By Irving Langmuir 
RESEARCH LABORATORY OF THE GENERAL ELECTRIC COMPANY. SCHENECTADY. N. Y. 
Communicated by A. A. Noyes. February 2. 1917 
In a recent paper^ I have developed a theory according to which 
all the forces involved in the structure of soHds and liquids, are similar 
in nature to the forces causing chemical combination. Thus conden- 
sation, evaporation, adsorption, cohesion, crystallization, liquefaction, 
viscosity, surface tension, etc., are manifestations of the forces already 
known to the chemist. In all these cases the range of the forces is 
limited to atomic dimensions except in so far as their effects may be 
transmitted from atom to atom. According to this theory, every atom 
in a solid or liquid is chemically combined to every adjacent atom. 
This chemical union may be strong or weak and may be characterized 
either by primary or secondary valence (Werner). 
In most inorganic solid or Hquid substances of the strongly polar 
type, the identity of the molecule is wholly lost, but in organic com- 
pounds the groups of atoms constituting the chemical molecule usually 
have a real existence even in the liquid or solid state. These group 
molecules are held together by primary valence forces while the forces 
acting between the group molecules, although no less chemical than 
the others, are to be characterized as secondary valence forces. 
From this viewpoint the forces involved in adsorption and surface 
tension do not originate from the group molecule as a whole, but rather 
from certain atoms in the molecule. 
This theory leads inevitably to the conclusion that adsorbed films on 
plane surfaces of solids or liquids should, in general, be one atom or 
group molecule in thickness. Considerable experimental evidence has 
already been presented that this is the case with films of gases adsorbed 
on solids. 
Miss A. Pockels^ showed, in 1891, that very small amounts of oil on 
the surface of water have no appreciable effect on the surface tension, 
