312 Thermo-Electricity and Electro-Magnetism. 



by taking two pieces of bismuth and antimony in the fingers and sim- 

 ply touching them ; 50°, 30° and 20° may also readily be obtained 

 by the contact of bismuth with the other metals under like circum- 

 stances. Examining some of my former results with this instrument, 

 I immediately discovered that I had assigned wrong places to anti- 

 mony, arsenic and gold. The two former must be included with 

 bismuth and other positive metals, while the gold falls in with the 



negative ones. 



Although the elementary conditions of the metals appear to be suf- 

 ficiently constant, some very singular results follow, when dissimilar 

 portions of metallic matter come into contact. I have shown that of- 

 tentimes the voltaic currents are reversed, by reversing the direction 

 of the caloric ; and subsequent research has furnished new and inter- 

 esting particulars. With my present multiplier, 1 find that antimony 

 and arsenic are capable of giving off both currents from contiguous 

 portions of the surface, when heated and touched by other metals. 

 These opposite electrical states seem to be connected with crystalli- 

 zation or irregularity of surface, I cannot say which, and they disap- 

 pear, at temperatures varying for each metal brought into contact. 



Thus, upon heating a bar of antimony and touching it, from time t& 

 time, as it cooled, by the other metals, I found the opposite currents 

 to disappear, with silver at a temperature near 280° F. — with gold, 

 near 90° — with lead at 82° — whereas, with bismuth, these opposing 

 currents might be traced even at the mean temperature. It is a most 

 singular fact that they may be drawn off by the galvanometer w r ires, 

 from portions of the antimony not more than \ of an inch apart. 

 They are constant, as to position, while they last. 



The next law, as it seems to me, is one of still greater interest. 

 In all the metallic combinations which I have yet examined, I have 

 observed that " the voltaic currents produced by the contact of dis- 

 similar metals, are the reverse of those occasioned by the friction of 

 the same metals" Some combinations, as lead and tin, will scarce- 

 ly produce currents under any circumstances, and bismuth and anti- 

 mony are most effectual, but the law may be observed even between 

 particles of the same metal, as bismuth. 



The two following tables are regarded as denoting with sufficient 

 accuracy, some of the thermo-electric relations. 



1st. When particles of the same metal, unequally heated, are 

 brought into contact. 



