PROCEEDINGS 
OF THE 
NATIONAL ACADEMY OF SCIENCES 
Volume 3 MARCH 15.1917 Number 3 
THE CONDENSATION AND EVAPORATION OF GAS MOLECULES 
By Irving Langmuir 
RESEARCH LABORATORY. GENERAL ELECTRIC COMPANY. SCHENECTADY, N. Y. 
Commmiicated by A. A. Noyes, January II, 1916 
Several years ago/ I gave evidence that atoms of tungsten, molyb- 
denum, or platinum vapors, striking a clean, dry glass surface in high 
vacuum, are condensed as solids at the first collision with the surface. 
Subsequently, similar evidence^ was obtained in connection with a study 
of chemical reactions in gases at low pressures. It was concluded that 
in general, when gas molecules strike a surface, the majority of them 
"do not rebound from the surface by elastic collisions, but are held by 
cohesive forces until they evaporate from the surface." In this way a 
theory of adsorption was developed^ which has been thoroughly con- 
firmed by later experiments. It was stated: ''The amount of material 
adsorbed depends on a kinetic equilibrium between the rate of con- 
densation and the rate of evaporation from the surface. Practically 
every molecule striking the surface condenses (independently of the 
temperature). The rate of evaporation depends on the temperature 
(van't Hoff's equation) and is proportional to the fraction of the sur- 
face covered by the adsorbed material." 
R. W. Wood* described some remarkable experiments in which a 
stream of mercury atoms impinges upon a plate of glass held at a defi- 
nite temperature. With the plate cooled by liquid air, all the mercury 
atoms condense on the plate, but at room temperature all the atoms 
appear to be diffusely reflected. 
The whole question of the evaporation, condensation, and possible 
reflection of gas molecules has been discussed at some length in two 
recent papers^ It was pointed out that, in Wood's experiments, 
there are excellent reasons for believing that the mercury vapor actually 
condenses on the glass at room temperature, but evaporates so rapidly 
HI 
