184 Mr. W. H. Patterson on 



The values of Haber and B rimer are marked by points ; 

 they are above the curve at low temperatures and beneath it 

 at high temperatures. 



The reason for the high values at low temperatures can be 

 understood, the molten alkali, according to Le Blanc and 

 Brode, being at this range of temperature hydroscopic. 

 The values at higher temperatures Haber and Bruner them- 

 selves regard as uncertain, their method seeming unfit to 

 yield more exact results. 



In order to obtain better values the author tried to find 

 electrodes which would, by insertion in molten alkali, behave 

 as hydrogen and as air-electrodes respectively, so that the 

 difference of potential between them would give at once the 

 E.M.F. of the oxy -hydrogen cell looked for. 



It seemed probable that palladium saturated with hydrogen 

 might be used as a hydrogen electrode. Palladium wires 

 were charged, either in dilute sulphuric acid or dilute alkali, 

 dried and inserted. Such wires develop hydrogen rapidly, 

 but the amount of gas cathodically incorporated with the 

 wire is so far above the amount which is in equilibrium with 

 hydrogen under atmospheric pressure, at the high tem- 

 perature of the molten alkali, that an excess of E.M.F. 

 amounting to some centivolts is brought about. Of course 

 the supercharge only lasts for a very short time, the hydrogen 

 being driven off rapidly and the E.M.F. falling slowly; 

 finally it reaches the value of the air electrode. There was 

 no marked break in this fall of potential at the value be- 

 longing to that in equilibrium with hydrogen at atmospheric 

 pressure. Hydrogenized palladium was therefore not suitable 

 as a hydrogen electrode. Iron was tried, a metal which, as 

 previously shown by Haber and Bruner, gives a value very 

 near to that of hydrogen gas. The difficulty which ensues 

 from the iron becoming too rapidly passive was overcome 

 by coating the iron wires used with electrolytic iron, after 

 the method of Classen, from a solution containing ferrous 

 oxalate or from a solution containing ammonium chloride. 

 Such wires show the hydrogen potential and for a longer 

 time. 



The possibilities with regard to the air electrode were 

 more numerous. Silver, platinum, palladium, nickel, and 

 iron were tried. Wires of these metals, after insertion into 

 molten alkali, all showed the air-potential — the iron and 

 nickel of course after some time — necessary to let them 

 become passive. There is possibly a slight deviation in the 

 case of palladium which, being probably slightly attacked by 



