332 C. G. KNOTT 



the hydrogenlum line was traced out by its neutral points with con- 

 veniently situated imaginary lines formed by comlnnation of iron, and 

 nickel wires with adjustable resistance in the two branches. The}'' 

 obtained for the point of conA'ergence the value —260° C. The con- 

 clusion seems to l^e that the hydrogen line, if it coidd be found, would 

 pass through this same point — /. e. have a neutral point with palla- 

 dium at temperature of about — 300° C. 



The thermoelectric peculiarities of hydrogenium may be prettily 

 shown by the following experiment. Let a palladium wire, by im- 

 mersion to half its lenfi^th in the electrolvtic cell, be hvdrogenised 

 throughout that half length. Attach the ends of this seeming single 

 uniform wire to the terminals of a galvanometer, and let a flame be 

 allowed to play gently at the central point of the wire. A large 

 current is at once obtained, which grows to a maximum, and then 

 diminishes to zero as the temperatiu-e rises to a red heat. There is 

 no such current during cooling. This spurious neutral point during 

 heating is due to the hydrogen being driven out of the heated portion, 

 partly, no doubt, into the contiguous colder portions, l'y following 

 up with the flame the ever-shifting point of separation between the 

 charged and uncharged portions, we may repeat the experiment 

 indefinitely until the hydrogen is all driven oat of the wire or until 

 the distribution of the charge has become fairly unif)rm. 



