ELECTRO-MAGNETISM. 



place in the directions of the current 

 when the metals contain any alloy, or 

 are not in a state of perfect parity. 

 Thus, although bismuth and tin are 

 each positive with regard to copper, yet 

 an alloy of the two former is found to 

 be negative with regard to the latter 

 metal. 



(317.) We are yet far from possessing 

 any theory by which the whole of the 

 facts belonging to thermo-electricity 

 can be satisfactorily explained. The 

 most intelligible account of them appears 

 to be that given by Becquerel*, which 

 proceeds upon the hypothesis, that when- 

 ever a particle of a metal receives heat 

 from a body of a higher temperature 

 than itself, part of the neutral electric 

 fluid which is attached to it is decom- 

 posed, the vitreous fluid being retained, 

 and the resinous fluid driven off, and 

 passing into the adjoining particles of 

 metal. In proportion as the heat ex- 

 tends, by communication from particle 

 to particle, similar effects take place in 

 each of those that are acquiring heat, 

 while contrary effects are taking place in 

 all those that are losing heat. Thus, the 

 simple diffusion of that portion of heat 

 which was originally received by the 

 first particle, produces only an oscilla- 

 tory movement of the electrical fluids 

 between adjacent particles, attended by 

 a series of decompositions and com- 

 binations of the two electric fluids. But 

 if the source of heat be permanent, so 

 that the temperature of the first particles 

 which receive it be uniformly maintained, 

 the retrograde movements of the decom- 

 posed electric fluids are prevented, and 

 a continued current of each takes place 

 in opposite directions ; the negative 

 electricity being impelled forwards from 

 the parts where the temperature con- 

 tinues high to those which continue to 

 be colder, and a positive current moving 

 in the contrary direction. It follows, 

 from this hypothesis, that when two 

 different metals are placed in contact, 

 so as to constitute a circuit, the currents 

 from the heated parts that are con- 

 joined will be urged in opposite direc- 

 tions ; but the strongest will prevail, 

 and the thermo-electric current actually 

 observed is that which results, and of 

 which the intensity is equal to the differ- 

 ence between the two that are simul- 

 taneously developed. 



(318.) Thermo-electricity does not 

 appear to have its source in any chemical 



* See Annales dc Cbimie ; tome xli. p. 353, 



changes taking place in the materials 

 composing the circuit. Oxidation, at 

 least, has no share in the effect ; for 

 Becquerel has repeated the experiments 

 of Seebeck and others, relating to this 

 mode of action, when the apparatus was 

 surrounded by hydrogen gas, without 

 any sensible difference in the results.* 



(319.) The great peculiarity which dis- 

 tinguishes thermo-electric currents from 

 those produced by galvanic action, is that 

 the quantity of circulating electricity is 

 much greater compared with its inten- 

 sity. They are, in this respect, still 

 further removed from the condition of 

 streams of electricity produced in the 

 common electrical machine, which pos- 

 sess amuchgreaterintensity.thoughthey 

 are much less considerable in quantity 

 than galvanic currents. Hence it is 

 chiefly by their effects in producing 

 deviations in the magnetic needle that 

 the existence of thermo-electric currents 

 is recognised. The low state of inten- 

 sity of these latter currents occasions 

 great loss of power whenever they have 

 to traverse any considerable line of con- 

 ductors, even of metals, which are the 

 most perfect conductors. On this ac- 

 count it is that very little advantage is 

 gained by forming compound circuits ; 

 that is, arranging their elements in a 

 series of alternations analogous to those 

 of the voltaic pile. Messrs. Fourier and 

 Oersted made trials of this kind ; first 

 combining three bars of bismuth with 

 three bars of antimony, placed alter- 

 nately, so as to form the sides of a 

 hexagon, and with their contiguous ends 

 soldered together, thus composing a 

 thermo-electric circuit, which included 

 three pair of elements. The length of 

 the bars was about 'four inches and a 

 half, their breadth about half an inch, 

 and their thickness one-sixth of an inch. 

 This circuit was placed upon two sup- 

 ports, and in a horizontal position, one 

 of the sides of the hexagon being in the 

 magnetic meridian. A compass-needle 

 was placed below this side, and as near 

 to it as possible, arid was very sensibly 

 affected when one of the solderings at 

 the junctions of the bars was heated 

 with the flame of a lamp. The devia- 

 tion was considerably increased on heat- 

 ing two of the alternate angles of the 

 hexagon ; and a still greater deviation 

 was produced when the heat was applied 

 to the three alternate angles. Similar 

 effects were produced when, instead of 



* Annales de Cbimie, tome xli, p. 359. . 



