330 



C. G. KNOTT 



Resiatauce. 



Temi)eiatiu-e- 

 Coefficient 



Prodnct. 



1 



1.07 



1.13 



1.27 



1.41 



1.51 



1.63 



.00293 

 .00319 

 .00299 

 .00273 

 .00257 

 .00214 

 .00198 



.00293 

 .00341 

 .00338 

 .00347 

 .00362 

 .00323 

 .00323 



When tlie peculiar conditions of the experiment are borne in 

 mind, more e.specially the uncertainty of the charge being uniformly 

 distributed in the Avire, it will probably be granted that the products 

 are sufi&ciently alike to justify the conclusion drawn. At any rate 

 the approximation to constancy among these products is very re- 

 markable when taken in connection with the rapid increase of 

 resistance due to charü^ing as shown in the first column. 



The thermoelectric current obtained Avith a palladium hydro- 

 genium pair is one of surprising magnitude, being greater than for a 

 palladium copper circuit. When the' heated junction is at or near a 

 temjjerature of 200° C, peculiar irregularities appear which depend 

 upon whether the temperature is for the moment rising or falling. 

 This no doubt is due to the hydrogen being driven out of the highly 

 heated parts near the junction, and to its partial return from colder 

 cimtiguous portions of the wire as the junction is being cooled down 

 again. After experiencing such a high temperature, the charged wire 

 at the junction does not, 4iOAvever, wholly regain its original con- 

 dition — the electromotive f^rce in the circuit for a given difference of 

 temperature being appreciably, sometimes very markedly, smaller than 

 at first. So long as the temperature is kept below 150° C, the 

 charged wire is as constant in its thermoelectric properties as the pure 



