Intelligence and Miscellaneous Articles, 295 



evident, and I should have deferred the publication of it until then, 

 had not Professor de la Rive advised me to join it to the preceding 

 obsei-vation, [Matteucci's Researches on the Torpedo, vide last and 

 present Numbers of Phil. Mag.] and ]to record its date in our So- 

 ciety. — Bibliotheque Universelle, November 1837. 



ON THERMO-ELECTRIC PH^INOMENA. BY CH. MATTEUCCl. 



Every time that a copper wire attached to a galvanometer and 

 well-brightened (decape) is brought into contact with a second cop- 

 per wire equally well cleaned, but heated by a lamp, we obtain an 

 electric current, which passes from the hot end to the cold end. If 

 we repeat this experiment with well-cleansed iron wires we obtain a 

 contrary current which proceeds from the cold end to the-hot one; 

 the same takes place also with zinc and antimony. This difference 

 is observed whatever be the temperature to which one of the wires 

 is heated. Now if instead of touching the wires we immerse them 

 in pure mercury contained in two capsules united by a siphon full 

 of mercury, one of which is hot, the other at the common tempera- 

 ture, we have still a current, but which proceeds in the same direc- 

 tion with copper, iron, zinc, and antimony. ITie mercury has here 

 no influence by the thermo-electric currents which it might deve- 

 lop, for the same results are obtained, if the two wires, always well- 

 brightened, are immersed one after the other in the same heated cap- 

 sule. It is therefore the action of the heat and of the air which pro- 

 duces an alteration at the surface of the metal : and in fact, if a 

 copper wire is heated exposed to the air in the flame of an alcohol 

 lamp, and afterwards immersed in the mercury where the other wire 

 is, the same difference is still observed as when the unequally heated 

 wires were placed one on the other. 



I endeavoured to obtain thermo-electric currents with mercury 

 by employing three capsules united by means of two siphons. In 

 the outer capsules were plunged the wires of the galvanometer. I 

 take away one of the siphons, heat the middle capsule and replace 

 the siphon. I thus bring the hot mercury into contact with the 

 cold. In this manner, however, I obtained but very feeble and doubtful 

 deviations. Although the wire of the galvanometer was rather 

 long, yet I doubt of there having been any development of thermo- 

 electric currents on the mercury. 



An amalgam of bismuth (1 of bismuth, J of mercury) which is 

 very crystalline, is endowed with a very considerable thermo- 

 electrical power. 



If a heated plate of bismuth is touched with the two extremities 

 of a galvanometer, very powerful currents are obtained. If the 

 bismuth be melted and we still retain the two ends immersed in the 

 fluid metal, the currents cease ; at times there are still some, but 

 we at once easily perceive that either some of the bismuth has so- 

 lidified, or that the two ends of the wire are unequally heated. With 

 a larger melted mass in any sort of a vessel these currents entirely 

 cease. If we then discontinue the heating, and allow the mass to 

 become cold, at the instant when it solidifies the needle indicates 



