hy Thermo-electric Currents. - 419 



two junctions of the bismuth and antimony element tlevelopes 

 an electromotive force, weaker in proportion as the difference 

 of temperature is greater, even between the limits 15° and 35°, 



This result is easily verified in the following manner. The 

 bismuth and antimony element being arranged in the two 

 vessels full of water, as in fig. 6, and in communication with 

 the galvanometer, the needle is brought to zero by raising the 

 water of the two vessels to exacdy the same temperature ; a 

 certain quantity of hot water is then poured into one of the 

 vessels, so as to produce precisely a difference of temperature 

 of 1° between the two solderings. The deflexion 7i of the 

 needle is noted down. 



After this a very feeble and perfecdy constant hydro-electric 

 current is passed through the second wire of the differential 

 galvanometer: the needle is deflected a certain quantity by 

 this current, but it is brought back to zero by suitably raising 

 the temperature of one of the bismuth and antimony junctions. 

 The two currents then hold each other in equilibrio. The 

 temperature of the same junction is raised 1°, whence results 

 a deflexion of the needle, which is precisely equal to the de- 

 viation n previously observed, if the electromotive force deve- 

 loped by an increase of 1° in the difference of temperature is 

 the same whatever that difference may be. 



By passing in this manner successively stronger and stronger 

 constant hydro-electric currents through the second wire of 

 the galvanometer, and neutralizing them each time by a suit- 

 able difference of temperature between the two solderings of 

 the thermo-electric current, I found that the electromotive 

 force developed by an increase of 1° of difference of temperar 

 ture was the more feeble as that difference was greater. The 

 thermo-electric element formed of wires of iron and platinum 

 is not the only one which I have experimented with at high 

 temperatures ; I likewise made some trials with elements com- 

 posed of other metallic wires. But the element of iron and 

 platinum has proved to be most suitable ; it is that the electro- 

 motive force of which decreases least with the elevation of the 

 temperature. 



The sensitiveness of a couple of iron and copper decreases 

 very rapidly with the temperature. At about 240° an eleva- 

 tion of 20° to 30° no longer exerts any influence upon the 

 needle, which remains perfectly stationary : the needle retro- 

 grades when the temperature is raised higher ; and the inten- 

 sity of the current, far from increasing with the temperature, 

 now decreases. This observation agrees with what M. Bec- 

 querel long ago observed on a couple of iron and copper. 

 According to that able philosopher, the current is even esta- 



