186 DR CARGILL G. KNOTT ON THE ELECTRICAL PROPERTIES, ETC. 



wire is hydrogenised throughout half its length, and its extremities attached to 

 the terminals of a galvanometer. If, now, a flame is applied to the centre of 

 this apparently single uniform wire, a large thermoelectric current is obtained, 

 which grows to a maximum and then falls down to zero. This spurious neutral 

 point is of course simply the result of hydrogen being lost to the heated portion. 

 As the wire cools down again no such large current is obtained. The effect 

 may be reproduced a number of times by following up with the flame the ever- 

 shifting point of separation of the hydrogenium and the palladium. 



Summary. 



The electrical resistance of hydrogenium increases with the temperature up 

 to the point at which hydrogen begins to be given off. Thereafter the resistance 

 begins to decrease till all the hydrogen has been driven away, after which 

 increase sets in again as the now pure palladium wire is heated. The tempera- 

 ture-coefficient diminishes as the charge increases, and in such a manner that 

 the total increase of resistance through a given range of temperature is nearly 

 the same for the same palladium wire whatever the charge may be. In other 

 words, at any given temperature below 150° C, the increase of resistance due 

 to a given additional charge is the same. The changes of resistance between 

 200° C. and 300° C. afford a very delicate means of studying the manner in 

 which the hydrogen escapes. 



The thermoelectric current in a palladium-hydrogenium circuit flows from 

 palladium to hydrogenium through the hot junction — that is, the hydrogenium 

 line lies higher in the thermoelectric diagram than the palladium line. The 

 higher the charge the higher the position. Saturated hydrogenium lies, for 

 ordinary atmospheric temperatures, between copper and iron. Up to 150° C. 

 the hydrogenium lines are straight lines, and nearly if not quite parallel to the 

 palladium line. Above 200° C. rapid changes set in, the result no doubt of the 

 loss of hydrogen at the junction. On cooling, the hydrogen seems to return 

 partially to the extremity from which it has been driven. The effects at high 

 temperatures are, however, complicated, because of the unequal distribution of 

 hydrogen in the palladium. 



