HYDROGEN INTO ATOMS 191 



to conclusions which could only be reconciled with the experimental data 

 by making many improbable and arbitrary assumptions. These difficulties 

 disappear, liowever, when we adopt the present theory of the mechanism 

 of the reaction. A brief discussion of the unsuccessful hypotheses will en- 

 able us to see that they are untenable. 



In attempting to analyze the probable mechanism of the reaction, three 

 possibilities suggest themselves : 



1. The dissociation occurs in the heated gas around the wire. 



2. The dissociation occurs on the surface of the wire by the impact 

 of the molecules and atoms against the surface. 



3. The dissociation occurs in the wire or in a surface layer where 

 equilibrium prevails. 



The first hypothesis assumes that the tungsten does not act catalytically 

 on the reaction, and is therefore in itself very improbable. Moreover, it 

 leads to conclusions irreconcilable with the experiments at low pressures, 

 where, in a sense, there is no heated gas around the wire. 



The second hypothesis seemed at first sight to have much in its favor, 

 and by its aid it was possible to estimate an upper and lower limit to the 

 degree of dissociation. 



Whatever the mechanism of the reaction on, or in, the wire, it was 

 possible to analyze the process as follows : 



Let 82 represent, out of all the hydrogen molecules which strike the 

 wire, that fraction which becomes dissociated into atoms. Similarly, £1 

 represents, out of all the atoms that strike the wire, that fraction which re- 

 combines to form molecules. We may consider that the reaction actually 

 pbserved is simply the difference between these opposite reactions. It is 

 clear that neither 81 nor 80 can ever exceed unity. 



For further progress, it is necessary to make some assumptions regard- 

 ing these coefficients. It is evident that either 81 or 82 or both must vary 

 with the temperature. The simplest assumption is that they are independent 

 of the pressure. By applying Equation 5 we then readily find that the 

 condition for equilibrium is 



p[ = V2 (esAi) pi- 



Since 81 cannot exceed unity, this gives us a lower limit for the degree 

 of dissociation, provided we can estimate 82. By assuming a value for the 

 heat of reaction it is possible to calculate 82 from Wd by an equation similar 

 to (16). In this way it was concluded that the degree of dissociation at 

 3200° K. and atmospheric pressure must exceed 5%.^^ 



*« Table XV gives 21%. 



