M. G. Wiedemann on the TVansmission of Heat in Metals. 399 



If from these numbers the tempei-atures ?/ and y^ of the bars 

 at their points of contact be calculated, we shall get — 



Copper-Bismuth 

 Zinc-Bismuth . 

 Copper-Iron I. 

 Copper-Iron II 

 Tin-Iron I. 

 Tin-Iron II. 

 Iron-Iron . 

 Iron- Copper 



The difference of temperature of two bars in contact is hence 

 very small. It is not greater in the case of two bars of different 

 metals, whether the heat pass from the better to the worse con- 

 ductor or the reverse, than in the case of two bars of the same 

 metal. 



Hence a resistance to the passage of heat through the place 

 of contact of two bars cannot be proved within the limits of 

 these observations*. 



If the two bars are not perfectly in contact, a decrease of tem- 

 perature at their place of contact is observed. The same happens 

 when the bars are separated by a thin layer of some bad con- 

 ductor. 



Experiments made in this manner gave the following series of 

 observations : — 



* For an accurate calculation of the relative conihictibilities of different 

 metals the numbers in question are not fitted, since the quotient (j of the 

 sums of the temperatures of two points divided by the temperature of the 

 middle point differs too little from 2. Hence a small error of observation 

 changes materially the calculated result. 



