1823.] Development of Elect romaghetism by Heat. 427 



Article V. 



On the Development of Electromagnetism by Heat. By the Rev. 

 J. dimming, MA. FRS. and Professor of Chemistry in the 

 University of Cambridge. 



(To the Editor of the Annals of Philosophy.) 



DEAR SIR, Cambridge, April 30, 1823. 



The following, selected from some experiments which were 

 the substance of a paper read at the last meeting of the Cam- 

 bridge Philosophical Society, as, I believe, they are new, will, 

 I hope, be found interesting to your readers. 



All the metals, not excepting fluid mercury, produced a devia- 

 tion on the magnetic needle of a galvanoscope, provided the 

 extremities of the metallic bars employed were at different tem- 

 peratures. 



It was indifferent whether the connecting wires were united 

 to the bars by soldering, rivetting, or simple contact. 



Bismuth and antimony (the metals which seemed most 

 efficacious) gave with copper wires, deviations in opposite 

 directions. 



A bar of bismuth four and a half inches long, half an inch 

 broad, and one-eighth thick, gave a positive deviation of 21° on 

 a compass needle, four and a half inches long, at the melting 

 point of bismuth ; at 180° and 100°, the deviations were respect- 

 ively 12° and 5°, the cooler end being constantly at 60°. 



A similar bar of antimony, with the utmost heat of a spirit- 

 lamp, gave a negative deviation of 19°. 



When a few drops of ether were made to evaporate from one 

 extremity of these bars, the other being kept at a constant tem- 

 perature of 60°, deviation was produced, but, of course, in the 

 opposite direction. 



A slip of palladium, weighing 35 grains, having silver wires 

 rivetted into it, when made red-hot at one extremity, gave with 

 a small compass needle a deviation of 70° positive ; withplatina 

 wires 10° ?iegative. 



A similar slip of platina with silver wires gave a positive 

 deviation of 65° ; with platina wires 4° negative. 



A bar of zinc cast upon copper wires gave a negative devia- 

 tion of 45°; the same bar with zinc wires gave a deviation of 

 2°; with silver wires a deviation of 2°; and with iron wires a 

 deviation of 3°, all positive ; but with platina wires a deviation 

 of 50° negative. 



A copper bar with zinc wires gave a deviation of 20° ; with 

 copper wires a deviation of 10° ; and with silver of 30°, all posi- 

 tive ; but with platina wires a deviation of 18° negative. 



A small silver bar with silver wires gave a positive deviation 

 of 20°; but with platina wires 60° negative, both at a red heat. 



