206 PHENOMENA, ATOMS, AND MOLECULES 



taken to be proportional to the corresponding pressures in the gaseous 

 phase. 



This theory led to an equation for the heat loss from the wire which 

 proved to be in excellent agreement with experiments extending over a 

 range of pressures from o.oi mm. up to 760 mm. and a range of tem- 

 peratures from 1000 to 3500° K. There are, however, some serious objec- 

 tions that may be raised against the assumption that the wire contains 

 concentrations of atomic and molecular hydrogen in equilibrium with each 

 other, and that these concentrations determine the rate at which hydrogen 

 is dissociated by the wire. We shall see, however, that the new theory en- 

 ables us to derive the same equation without making these objectionable 

 assumptions. 



The hypothesis that solid solutions of gases in metals play an important 

 role in chemical reactions in contact with metals is a common one, and 

 usually is not looked upon unfavorably. But it is evident that if we are to 

 retain our ordinarv conception of concentration the reaction must take 

 place in a volume which contains many layers of atoms. In the experiments 

 the range of concentrations of molecular (or atomic) hydrogen in the 

 metal phase must have been enormous. Thus under some of the experi- 

 mental conditions the concentrations were so small that the individual 

 atoms or molecules of hydrogen were certainly separated from each other 

 by hundreds of tungsten atoms. The rapidity of the reaction under such 

 conditions could not possibly be very high and would be limited by the rate 

 of diffusion of the gas through the metal. 



Yet the experiments showed clearly that the velocity with which 

 equilibrium was reached on the surface was extraordinarily high under 

 all conditions. In fact, it was so high that practically all of the hydrogen 

 atoms and 68% of the molecules which struck the surface reached equilib- 

 rium before leaving it again. When we consider that at atmospheric 

 pressure and room temperature i.i X lo^"* molecules of hydrogen strike 

 each square centimeter of surface per second, and that this number of 

 molecules is contained in about 44 liters of gas, we realize the enormous 

 velocity which this reaction must have. It is certainly impossible that such 

 amounts of gas per second could diffuse into any kind of solid body and thus 

 be brought to equilibrium. 



These considerations compel us to assume that the reaction occurs 

 directly on the surface of the metal and that it does not involve the 

 diffusion through a film even as thin as that of a single layer of atoms. 



A second difficulty arose in connection with the original theory. 



At temperatures up to about 1500° K. the accommodation coefficient 

 of hydrogen in contact with tungsten was found to be equal to 0.19. In 

 other words, only about 19% of all the hydrogen molecules striking the 



