HYDROGEN INTO ATOMS 207 



filament, reached thermal equilibrium with it before leaving it. This figure 

 was in good accord with the value 0.26 obtained by Knudsen with platinum 

 at room temperature, especially when the fact was taken into account that 

 accommodation coefficients in general appear to have slight negative tem- 

 perature coefficients. 



On the other hand, at high temperatures, it was found that 68% of all 

 the hydrogen molecules striking the filament reached chemical equilibrium 

 before leaving it. The original theory gave no clue to the solution of the 

 paradox that 68% of the molecules reached chemical equilibrium, while 

 only 19% reached thermal equilibrium. It is certainly impossible that 

 molecules should reach chemical equilibrium while great differences in tem- 

 perature still persist. 



The viewpoint which has served as the basis for the new theory was 

 gradually developed in quite other lines of work, particularly in connec- 

 tion with a study of the effect of gases on the electron emission from 

 heated metals.^ It was found that the electron emission was dependent on 

 the composition of the actual surface layer of atoms. The deeper layers 

 were apparently without effect. Subsequent work on gas reactions at very 

 low pressures ■'^ indicated clearl}^ that chemical reactions between a gas 

 and a solid depend in general on the composition or structure of the outside 

 layer of atoms of the solid, rather than on the thickness of an adsorbed 

 film through which the gases diffuse. 



This theory has now been developed much more completely than in any 

 of the previously published work. 



It is definitely known from the work of Bragg and others that the 

 atoms of crystals are arranged according to space lattices in such a way 

 that the identity of molecules is usually lost. The forces holding the 

 crystal together are thus clearly chemical forces which act probably only 

 between adjacent atoms. On the surface of a crystal the forces tend to be 

 chemically unsaturated, and atoms or molecules of gases can thus be held 

 firmly by the atoms of the solid.'' In general the law of multiple com- 

 bining proportions will apply. Thus each metal atom of the surface will be 

 capable of holding a definite integral number (such as one or two) of atoms 

 of the gas, or possibly each two atoms of metal may hold one atom of gas. 

 The atoms held on the surface in this way will form a part of the solid 



^ Langmuir, Phys. Rev., 2, 450 (1913) ; Physik. Z., 15, 516 (1914). 



^ Jour. Amer. Chem. Soc, 37, 1139 (iQiS)- 



^Haber {Jour. Soc. Chem. Ind., .?j. 50 (1914) and Z. Elckfrochem., 20, 521 

 (1914)) lias suggested on the basis of Bragg's theory, that adsorption may be the 

 result of unsaturated chemical forces at the surface of a solid body. Haber, however, 

 only considers the force causing adsorption and does not take into account that the 

 amount of adsorption depends on a kinetic equilibrium between the condensation and 

 the evaporation of molecules. 



