hy Thermo-electric Cwrenis, 



415 



the two thermo-electric elements remain comparable, it will 

 suffice once for all to form a table, in which are inscribed on 

 one side the differences of temperature, T' — ^, T"— /?, T"'—t, 

 of the iron and platinum element measured upon an air-ther- 

 mometer, and on the other side the differences of temperature, 

 Q'—Q, 5"— 5, Q"' — Q, of the bismuth and antimony element. 



If it is now desired to measure a high temperature with the 

 iron and platinum couple, it will suffice to look out the tem- 

 perature, 6' — 5, which equilibrates it on the couple of bismuth 

 and antimony ; and we shall find in the table, the construction 

 of which I have described, the temperature T—i correspond- 

 ing to it on the element of iron and platinum. 



This method is entirely independent of the measuring ap- 

 paratus. The magnetic state of the needle may vary without 

 any inconvenience resulting, as it will be modified in the same 

 manner for the two thermo-electric elements. The sole indis- 

 pensable condition is, that the two elements should always 

 remain perfectly comparable, and experiment will readily de- 

 cide whether ihis condition is satisfied. 



I shall not transcribe here the numerous experiments which 

 I have made by this method, but shall content myself with re- 

 lating in detail one of the series, and shall merely give the 

 results of some others in order that the pi'ogress of the obser- 

 vations may be judged of. 



