270 Mr. Prideaux's Experimental Contributions 



36. A staple was then made, having one leg a tube of tin 

 plate, in which short cylinders of bismuth and tin were made 

 to slide. When either of them approached a heated point in 

 the tube, a current was driven back toward the heated point. 

 Thus it acts inside a tube as well as outside a wire. 



37. It is therefore evident that a foreign metal, brought into 

 contact with a homogeneous circuit, near a heated point, enters 

 into the thermo-electric connexion, and determines a current, 

 according to the reaction between the two metals. Such cur- 

 rent is, however, weaker than when the second metal inter- 

 cepts or forms a distinct portion of the circuit, and is what 

 might be expected, regarding the latter only as participating, 

 with the homogeneous metal, in the calorific action on the 

 portion to which it is applied. 



38. BecquerePs magnetest was much more delicate than 

 mine, and his results (13.), p. 213, coincide with the property 

 of therrno-negative metals (21.). I have applied this mode of 

 examination to zinc, heating a staple of polished zinc sheet at 

 the bend, and pinching it with another slip from the same 

 sheet, at different distances from the point of heat, conse- 

 quently at different temperatures. The effects varied with the 

 temperature, and also, apparently, with the direction in which 

 heat was flowing ; but they were not decided enough to par- 

 ticularize here. 



VIII. In what consists the superior efficacy of tinned surfaces ? 



39. It was noticed above (9.) that iron, instead of having 

 its thermo-electric action weakened by tinning the surface, 

 seemed to have it considerably augmented. To place this out 

 of doubt, a staple of iron wire had two loops turned out near 



the bend ; of these one was tinned, the other left clean, and 

 polished bright. The points also being tinned and connected 

 with the magnetest, the bend was heated nearly to redness, 

 and the loops quickly became hot. The bright loop was then 

 pinched successively with copper, tin, lead, and platinum, but 

 without much effect on the needle. The same metals were 

 then applied to the tinned loop, when the tin gave a deflection 

 of 10°, and the others in proportion. 



40. A platinum wire, 12 Xy^th inches, was then bent into 

 a staple, and at \ inch from the bend a bright iron wire, y 1 -^ in 

 diameter, was wound 16 turns round it, so as to inclose the 

 platinum in an iron helix for an inch in length. At \ inch on 



