towards the Theory of Thermo-electricity. 271 



the other side of the bend was wound a helix of tinned iron wire; 

 the dimensions of the wire and of the helix being the same as 

 the clean iron one. On heating the bend, a deflection of 10° 

 took place, the current setting towards the tinned helix. 



41. Equal lengths of tinned and untinned iron wire, of the 

 same dimensions, were twisted together at one end, so as to 

 form a pair. No deflection ensued on heating the twist; but 

 turning the tinned leg round a knob of bismuth, as above (36.), 

 the needle deviated, on heating the twist, 15° ; whilst the same 

 knob of bismuth, similarly bound in the clean leg, gave with 

 the same heat only 4° deflection. 



42. An iron wire was then cut into lengths of 16 inches, and 

 bent into staples, which were coated as follows: — 1, with tin; 

 2, bismuth; 3, lead; 4, fusible metal, all by heat; 5, with 

 copper, by the humid process. They were heated successively 

 at the bend to 140°, and the forceps (so we may denominate 

 the doubled slip of metal employed to pinch them, as above 

 (32. 33.),) applied at \ inch from the point of heat. The re- 

 sults are expressed in the following table. 



Forceps. 



These numbers are evidently not precise ; but, being the 

 means of several experiments, are indications sufficient for the 

 present purpose. 



The sign + signifies that the coated wire was + to the for- 

 ceps, i. e. that the current set from the point of contact toward 

 the point of heat : the sign — indicates, of course, the converse 

 direction. The copper coating differed little from naked iron ; 

 the tin exalted the thermo-electric power of the wire ; the fu- 

 sible metal did the same, in an inferior degree ; the lead very 

 slightly, if at all ; but the bismuth reduced it. At the foot of the 

 last column, iron coated with bismuth was — to bright iron. 



43. Similar staples of the same iron wire were bound tightly 

 with lead and tin foils, in single, double, and triple thicknesses. 

 But in every case, and with whatever forceps, the wires so 

 bound were inferior in action to clean iron wires. 



The efficacy of copper wire was also increased by tinning 

 and by amalgamation, but in a less degree than that of iron. 



44. I am inclined to attribute the efficacy of tinning to the 

 presentation of a soft and clean metallic surface, by which good 



