XXIV.] THE ELECTRO-MOTIVK l'<>\\ Kit OF HEAT. 337 



A is the section of a semieylindrical bar of antimony, B 

 of one of bismuth, united together by the opposite cor- 

 iHT.s of a lozenge -shaped piece of copper, C. From its 

 exposing so much surface, the copper becomes hot and 

 cold with the greatest promptitude, and from its good 

 conducting power it may be made very thin without 

 injury to the current. With a pair of bars three fourths 

 of an inch thick, and a circular cop- 

 per plate, as at D, Fig. 58, having both jfSHjS^I&tL 

 surfaces blackened, I have repeated A 

 the greater part of those experiments - '\_^ 



which M. Melloni made with his mul- 

 tiplier. A platinum wire, as at E, Fig. 59, may sometimes 



be very advantageously used. 



4. The currents circulating in a steel 



magnet are to all appearance perpetual. 

 pln , og I thought for some time it might be 



possible to procure similar perpetual 

 currents by compound thermo- electric arrangements. 

 Let a, b, c be wires of three different metals soldered to- 

 gether so as to form a triangle. Now if these metals 

 were selected, so that a and b could form a more power- 

 ful thermo-electric pair than a and c, or b and 6', it might 

 be expected that at all temperatures an incessant current 

 would run round the system. Such, however, will not 

 be found to be the case. In effect, any one of these three 

 serves simply as a connecting solder to the other two, 

 and hence no current is excited ; for the ends that have 

 the third metal between them, although that metal inter- 

 venes, are under exactly the same condition as the other 

 ends which are in contact. 



Y 



