396 M. G. Wiedemann on the Transmission of Heat in Metals. 



At the place of junction a slight displacement of the thermo- 

 electric element (with the exception of the case of the two last 

 bars) causes a considerable deflection in the galvanometer. 



In all cases the temperatures of the colder bars form almost a 

 geometrical series^ and the bars can hence be considered infi- 

 nitely long. 



If heat passes from one metallic rod into another equally thick 

 but infinitely long, the two rods must show a change of tempe- 

 rature at their junction if the heat experience a resistance in its 

 passage*. 



If the temperature y, at the place of contact of the colder bar 

 be calculated by multiplying the value under — 1 with the mean 

 ratio of the geometrical series which the temperatures of the 

 colder bars form; if, further, this value be put equal 100, and 

 the other temperatures of the bars calculated accordingly, we 

 can, by an accurate deUneation of the curve of heat corresponding 

 to the warmer bar, find the value y of the temperature of the 

 warmer bar at the place of contact. 



The mean of the present and other series of experiments is 

 found to be, — 



* Vide Poisson's TMorie maihematique de la Clialeur. If y and y, be 

 the temperatures of the warmer aud colder bars at tbeir place of contact, 

 and if r be a constant dependent upon the resistance to passage, and differ- 

 ent for diflfereut metals, the quantity of heat, w, which passes in a unit of 

 time through the point of contact is 



w=-{y-y{). 

 r 



If the difference of temperature y — yi=0, the constant of the resistance 

 of passage r must also be equal to 0, for otherwise the quantity of beat 

 passing tln-ough the j)oint of contact would be w=0. 



