towards the Theory of Thermo-electricity. 269 



Having placed the thermo-electric pair described (9.) p. 21 1, 

 in connexion with the magnetest, heat was applied to the mid- 

 dle of the copper leg instead of the twist. This soon commu- 

 nicated, through the good conducting power of the copper, to 

 the point of contact, and proportionate deflection ensued. 

 When the apparatus had cooled, heat was applied to the mid- 

 dle of the iron leg, which did not soon affect the needle. In 

 the meanwhile it was touched, near the heated point, with a 

 piece of iron ; then of tinned iron ; but still with little effect 

 on the magnetest. Presently the heat was renewed, and a 

 slip of sheet copper substituted for the iron, as a better con- 

 ductor, being also bent double, to pinch the wire, and thus 

 ensure good contact. A lively deflection now took place, but 

 in direction contrary to that above stated, the current being 

 from the part touched with copper to the point of heat. 



33. Instead of the pair, a single wire of tinned iron was 

 now taken, 18 by T ^th inches, bent in the form of the letter U, 

 or forming a sort of wire staple. The points of this being set 

 in communication with the magnetest, and the bend, or any 

 part (distant from the mercury*) heated, gave a current, when 

 pinched by copper, near the point of heat, from the part touched 

 to the heated point. 



34. A copper staple was now employed, and being pinched 

 with tin, gave the same result; but tin plate being used, in- 

 stead of tin, gave a current in the opposite direction, i. e. from 

 the heated point toward the part touched. Two flakes of an- 

 timony, stuck on the ends of a slip of card, produced the same 

 effect as tin plate. 



Thus foreign metals, applied to the surface of the wire, af- 

 fect the circuit. Is the thermo-electric action, then, super- 

 ficial ? 



35. A copper staple was taken, one leg of which took three 

 turns round a knob of antimony. Being connected with the 

 magnetest, heat was applied to the copper, just by the anti- 

 mony : strong deflection took place, as if the antimony had 

 formed part of the circuit, the current flowing from the heated 

 point to the part in contact with the antimony, and this to 

 whichever side the heat was applied. Bismuth being substi- 

 tuted for the antimony, currents took place equally active, but 

 in the converse direction ; and the same with tin, when sub- 

 stituted for the bismuth. Hence the inside of a spiral is 

 equally thermo-electric with the outside. 



* Liquid mercury is an active thermo-electric ; and care must be taken, 

 wherever it is used for connexion, not to allow any heat to reach the point 

 of contact with it. Many precautions are sometimes requisite to prevent 

 this. 



