and Conducting Power of Solid Bodies. 165 



of the multiplier seldom returned to precisely the point from 

 which it had been deflected, sometimes becoming stationary a 

 little to the right, sometimes to the left of its original position, 

 the cause of which deviation could not be traced, would give 

 rise to a slight error, perhaps O'l to 02 of a degree, upon 

 the scale of the multiplier, which had been divided in the 

 manner recommended by Melloni. 



The divisions upon this scale not corresponding exactly with 

 the temperatures they are intended to indicate, might also 

 lead to a slight error ; and although the observations were 

 always made with the aid of a magnifier, still it is possible 

 that from 0*1 to 0*2 of a degree should escape observation. 



These three sources of error, which arose chiefly from the 

 author's having used an imperfect instrument, added to another 

 more important one, the effect, namely, of currents of air in the 

 room cooling one part of the rod under examination more 

 than another, the author calculates will not amount to more 

 than O'^ C. in the final result of any experiment. In fact it 

 never actually amounted to so much. 



The metals used in the experiments were copper, steel, tin 

 and lead ; they were all drawn out into cylindrical wires or 

 thin rods, and their length was such that even in the middle 

 of the rod no effect of the heating body Was perceptible. The 

 copper alone, being the best conductor, was slightly affected 

 through its whole length. 



The object of the experiments was not so much to determine 

 the conducting power of the metals employed, as to submit 

 the analytical law to the test of experiment ; the metallic sur- 

 face of the rods was therefore not protected, and remained un- 

 impaired in the three first metals ; the lead wire however soon 

 became covered with a layer of oxide, which increased in 

 thickness every time it was heated. 



The results at which the author arrived were principally 

 the following: — 



The law of Biot, — that in a very thin, long metallic rod, 

 one end of which is kept at a constant temperature above that 

 of the surrounding air, after equilibrium of temperature has 

 been established, the excess of temperature in any part of the 

 rod above that of the surrounding air, decreases in geometrical 

 progression as the part examined is removed by equal distances 

 from the heated end — is not generally confirmed by the author's 

 experiments, and is only true for most of the metals in the 

 case of a very small excess of temperature. Among the metals 

 examined copper was the only one for which the law held good, 

 at least when the excess of temperature amounted to 30 C. 



With tin, the law no longer applied when the excess was 



