252 
PHYSICS: 1. LANGMUIR 
but that the surface tension begins to decrease suddenly when the 
amount of oil per unit area is increased beyond a certain sharp limit. 
Lord Rayleigh,^ Devaux,^ and Marcelin^ have made similar experi- 
ments and have concluded that the amount of oil needed to produce 
an appreciable effect on the surface tension corresponds to a layer one 
molecule deep. Devaux finds with triolein that a film 11. X 10-^ cm. 
thick just begins to lower the surface tension of water. From the den- 
sity and molecular weight of the oil and from Perrin's value of the 
Avogadro constant, he calculates that the diameter of the molecule 
would be 11.3 X 10~^ cm. if this be assumed spherical in shape. 
According to my theory, however, molecules should not be regarded 
as spheres, since such a supposition would not be consistent with the 
chemical nature of the forces. The fundamental question immediately 
arises : What causes the spreading of an oil upon the surface of water? 
If we regard molecules as spheres, any attraction between the water 
and the oil should cause the oil to dissolve in the water instead of 
spreading on the surface. From the chemical viewpoint, however, the 
force causing the spreading should be a force between atoms, not be- 
tween molecules. Evidently, then, some atom or atoms in the oil 
must have an afiiQity for the water. Now it is known that the pres- 
ence of the — COOH, —CO, or —OH groups in an organic molecule 
tends to render the substance soluble in water, while the hydrocarbon 
chain decreases the solubility. On the other hand, hydrocarbons are 
soluble in each other. Therefore the — COOH, —CO, and —OH groups 
have more affinity for water than for hydrocarbons, while hydrocar- 
bons have more affinity for each other than for water. 
Thus, when an oil is placed on water, the —COO— groups combine 
with the water, while the hydrocarbon chains remain combined (sec- 
ondary valence, of course) with each other. This process leads directly 
to the spreading of the oil on the surface. If only a limited amount of 
oil is placed on a large surface, the spreading ceases as soon as all the 
available —COO— groups have come into contact with the water, for 
any further spreading would separate the hydrocarbon chains from each 
other. ' 
According to this conception, pure hydrocarbon oils would no-t spread 
on water. Experiment shows that this is actually the case.^ 
The theory furthermore indicates that the hydrocarbon chains in 
the oils must be placed above the —COO— groups on the surface. 
Thus, in a series of homologous compounds, as the length of the chain 
increases the thickness of the oil film should increase in proportion, 
while the area occupied by each molecule should remain constant. 
