Hydrogen ivith Oxygen at Loiv Pressures, 473 



ions from its surface. This phenomenon has been the subject 

 of several recent researches. Mr. G. Owen (Phil. Mag. 

 Sept. 1903) investigated it by Mr. C. T. P. Wilson's method 

 of producing condensation-nuclei through sudden expansions. 

 By this means, he detected the discharge of negative ions 

 from platinum at 160° centigrade. The discharge was per- 

 ceptibly greater when the temperature reached 300°, and 

 became far more copious when the platinum was heated to 

 400°. Again, I have shown that a proportionate relation 

 exists between the number of molecules of water formed 

 during an electric discharge in rarefied hydrogen and oxygen 

 and the number of ions which pass to the electrodes, so long 

 as the circumstances of the discharge remain unaltered 

 (Phil. Mag. Feb.1904, Jan.1905). I suggested in explanation 

 of experimental results, that the molecules of hydrogen and 

 oxygen are dissociated into atoms by the impacts of the ions 

 moving swiftly under the electric force, and that the liberated 

 atoms group themselves anew into molecules of water. Is it 

 not more than possible, in view of Mr. Owen's experiments, 

 that such also is the explanation of the phenomena described 

 in this paper ? That when the platinum is heated to a 

 definite temperature, about 280° C, corpuscles are discharged 

 from it with sufficiently high velocities to dissociate the 

 molecules which they strike, and are discharged in sufficient 

 numbers to reveal the chemical change following upon this 

 dissociation. 



If this is the process at work, a certain amount of chemical 

 action is doubtless produced by the platinum when its tem- 

 perature is far lower than that at which the action is 

 perceptible ; for even then, according to Mr. Owen, the 

 platinum is discharging ions. But if the chemical action 

 does not absolutely begin when it becomes apparent, it must 

 at least then undergo an abrupt increase. 



It therefore becomes necessary to suppose that the corpus- 

 cular discharge from platinum makes an abrupt change in 

 extent or character — a change, e. g., in the number or sign 

 of the ions; or in the distribution of their velocities — at the 

 moment when a slight increase in the temperature of the 

 platinum carries the metal up to that critical temperature at 

 which the catalysis is first perceived. 



Connected with this temperature is another circumstance, 

 which has been mentioned above. After the platinum had 

 been heated up to this temperature and a little water vapour 

 had been formed, the chemical action could be afterwards 

 renewed by heating the platinum up to a temperature distinctly 

 lower than that which was needed to initiate the action. This 



