towards the Theo?y of Tliermo-electricity. 263 



inches, and the point amalgamated. Thus the faces of the 

 spirals could be pressed against each other, by means of the 

 cork handles, the amalgamated ends entering the mercury 

 boxes to communicate with the magnetest. One of the spirals 

 being now heated, and (having placed the points in the mer- 

 cury boxes,) pressed against the cold one, a deflection took 

 place of 5°, showing an electric current from the hot spiral to 

 the cold one. The heat employed was about 300°. 



20. An iron pair, formed in the same way, being similarly 

 employed, gave a deflection in the opposite direction, indica- 

 ting the currents/row the cold fothe hot spiral, which increased 

 by increased heat till, on approaching redness, it reached 18°. 



A pair, of tinned iron, being next employed gave the same 

 result, but in a much greater degree. The mere warmth of 

 the fingers communicated to one of the spirals produced de- 

 flections of 3° or 4<°; and before the heat reached redness, it 

 was 30° ; the electric current being in this instance also from 

 cold to hot. 



21. Platinum, silver, lead, tin, and bismuth all corresponded 

 with copper, in the currents passing from the hot to the cold 

 metal. Bismuth could not, of course, be obtained in spirals ; 

 and therefore rods of the other metals were employed to com- 

 pare with it. This property, in which the cold metal is posi- 

 tive to the hot (in the common thermo-electric sense*), may 

 be called ther mo-negative. Bismuth possesses it in the highest 

 degree; then silver; next platinum, copper, tin. Lead, whe- 

 ther in wire or in bars, has only just enough of it to recognise. 



22. Zinc and antimony correspond with iron, or are thermo- 

 positive ; that is, the hot metal is positive to the cold. But 

 the experimenter will have some trouble, at first, in ascertain- 

 ing the point. 



Rolled zinc gives a current from hot to cold, when heated 

 up to about 200°; between that and 250° is a neutral point; 

 and from 400° up to the melting point the current is more 

 lively, but reversed ; i. e. from cold to hot, or thermo-positive. 

 I have hitherto had some difficulty as to the temperature at 

 which the change takes place. When heated up to the 

 thermo-positive point, it seems to retain that character in 

 cooling, down to a temperature considerably below that at 

 which it first assumes it; and the direction of the heat, in the 

 heated bar, seems to exercise an influence in the effect. This 

 property of zinc explains the discrepancy respecting it in the 

 tables of Becquerel and Gumming; the former having worked 



* It must be noticed that the terms « positive' and c negative' have been 

 fixed in thermo-electricity in the sense of electrics, not of conductors ; po- 

 sitive here being that which takes, not that which gives, positive electricity. 



