produced by changes of Temperature . 307 



currents, and the deviation will then correspond to a force 

 triple of that which we would have had if we had only consi- 

 dered a single current. In causing to vary equally the quan- 

 tity of electricity which passes into each wire, it becomes easy 

 to compare the deviations of the needle with the correspond- 

 ing intensities of the electric current. 



Nothing is more easy than to procure equal currents. It is 

 sufficient to solder to each of the ends of the same wire one of 

 the extremities of an iron wire, so as to form three closed cir- 

 cuits, then to bend each of them to each soldered joint simi- 

 larly placed, in order to pass the curved part into a tube of 

 glass, closed at one of its extremities, and plunged into a mer- 

 curial bath, whose temperature is raised by a spirit lamp. A 

 thermometer also placed in the mercury indicates the changes 

 of temperature. In proportion as it is heated, the magnetised 

 needle, according to the discovery of M. Seebeck, deviates 

 from its position of equilibrium, and if we submit successively 

 to experiment one soldered joint, two soldered joints, and mark 

 at each time the deviation of the magnetised needle for the 

 same temperature, we shall have the angle which corresponds 

 to single, double, and triple forces. 



This method of experimenting requires great precautions 

 if comparable results are required. We must first plunge the 

 soldered joint, whose temperature is not to be raised, into melt- 

 ing ice, and the thickness of the tube into which is passed 

 part of the circuit, where the soldered joint is, ought to be 

 sensibly the same as that of the thermometer, in order that 

 the mercury of this instrument, and the metal of the circuit, 

 may receive the heat in the same time. Experiment proves al- 

 so that a mercurial bath is preferable to one of oil, on ac- 

 count of the great difference of conductibility of heat between 

 oil and the metal, a difference which occasions retardations in 

 the simultaneous production of phenomena. It is also neces- 

 sary, that, in the curved part of the circuit, whose temperature 

 is raised, the metals have no other points of communication 

 but those of their contact, for the intensity of the current 

 would certainly experience modifications from such a cause. 

 This inconvenience is avoided l>y covering with silk one of the 

 wires; and care must be taken thai the thermometer and tin 

 VOL. v. NO. II. OCTOBER 1 K2(). X 



