ELECTRO-MAGNETISM. 127 



that assigned to them in Volta's theory of contact. He also observed that cur- 

 rents were produced by inequality of temperature in bars of a single metal, 

 when they have a crystalline structure ; and suggested that the changes of 

 temperature of the metallic nucleus supposed by Davy to be within the exter- 

 nal crust of the earth, might produce those currents circulating round the 

 globe to the influence of which Ampere ascribed the magnetism of the globe. 



In the year 1823, these inquiries were prosecuted by the chevalier Yelin, 

 and MM. Marsh and Gumming.* The first investigated the influence of the 

 nature and form of homogeneous metals on the direction and intensity of the 

 electric current. The two latter philosophers produced the revolution of ther- 

 mo-electric currents round magnets. Soon afterward, MM. Oersted and Fou- 

 rier communicated to the Academy of Sciences a series of experiments on 

 currents obtained by thermo-electric piles, made by combining bars of anti- 

 mony and bismuth in a series. The alternate points of junction were heated, 

 and the current was manifested by the deflection of a magnetic needle. This 

 deflection, though considerable, was not observed to increase in proportion to 

 the number of thermo-electric elements constituting the pile. They attempted, 

 without success, to effect chemical decompositions by the current thus ob- 

 tained. This has, however, been since effected by Becquerel, by exposing to 

 the action of the thermo-electric current solutions easily decomposable. M. 

 Bottot, of Turin has also succeeded in decomposing water, and various so- 

 lutions, by the thermo-electric current obtained from a pile formed of a series 

 of wires of platinum and iron. 



The result of these and subsequent investigations of Seebeck, Becquerel, 

 and others, has led to the establishment of the following series of metals pos- 

 sessing thermo-electric excitability, in the order in which they stand : 



1. Bismuth. 5. Tin. 9. Zinc. 



2. Platinum. 6. Gold. 10. Iron. 



3. Mercury. 7. Silver. 11. Antimony. 



4. Lead. 8. Copper. 



If a thermo-electric couple be formed by any two metals in this series, the 

 positive electricity at the heated point will pass from that metal which holds 

 the higher to that which holds the lower place in the series ; consequently, 

 each of the metals in the series is thermo-electrically positive to all above it, 

 and negative to all below it. 



The intensity of many thermo-electric currents increases in proportion to 

 the temperature up to 40 R., but not after ; and at a certain point it falls. It 

 appears, too, from the experiments of M. Becquerel, that each metal has for 

 itself a proper thermo-electric power, which is the same for any circuit. He 

 thus expresses it : 



Metals Thermo-electric power. 



P. Iron. 5 



P. Silver 4-JJ7 



P. Gold 4-Oo2 



P. Zinc 4-035 



P. Copper 4 



P. Tin 3-1 



P. Platinum 3 * 68 



M. Nobili obtained thermo-electric currents by the contact of a hot and a 

 cold cylinder of porcelain, on each of which was moist cotton. M. Becquerel 

 considers that the water, at two temperatures, is here the exciting cause. The 

 rank of the chief metals, in the thermometric series, beginning with the posi- 



* BibL Univ., torn, xxiv., xxv., xxvii., and xxix. 



