Table 282. 

 RELATION BETWEEN THERMAL AND ELECTRICAL CONDUCTIVITIES. 



That there is a close relation 

 between the thermal and the 

 electrical conductivities of 

 metal was shown experimen- 

 tally by Wiedemann and Kranz 

 in 1S53, and had been referred 

 to by Korbes, with whom a 

 difficulty arose with regard to 

 the direction of the variation 

 with temperature. The ex- 

 periments of Tait and his stu- 

 dents have shown that this 

 difticulty was largely, if not 

 entirely, due to experimental 

 error. The same relation has 

 been sliown to hold for alloys 

 by Chandler Roberts and by 

 Neumann. This relation was 



a. Values in Arbitrary Units at 15^ C. 



denied by H. F. Weber, and 

 has been apain experimentally 

 iiivestif;ated and apparei.ily 

 established by the experiments 

 of Kirchhoff and Hanscmann, 

 of L. Lorenz, of 1''. Kohl- 

 rausch, and of I5erget. 



Putting /=z thermal conduc- 

 tivity, and /• = electrical con- 

 ductivity, Kirchhoff aiid 

 Hanscmann find the values in 

 Table a. This table shows 

 iron to deviate consideiably 

 from the other metals in the 

 relationship of the two con- 

 ductivities ; but this may possi- 

 bly be explained by its mag- 

 netic properties. 



Lorenz's results • show that the ratio // i for the different metals, except iron, is nearly constant for values 

 at o'-"' and lOo-"" C, but that the ratio is generally greater for poorly conducting substances. He shows that the 



llil2.^ -^ remains nearly constant for all metals examined, with the exception of iron, and has an aver- 



/t-i 



ratio ^ 



age value", as shown by Table b, of about 1.37. He concludes that i/k— constant X T, where T is the abso- 

 lute temperature. . . , , , j • .1. j t 

 In this table the values of / and i are given in c. g. s. units, and the metals are arranged in the order ot 

 their heat conductivities. The same specimens were used for both the thermal and the electrical experiments. 



b. Values in C. G. S. Units. 



Substances. 



Copper 

 Magnesium 

 Aluminium 

 Brass, red . 

 Cadmium . 

 Brass, yellow 

 Iron . 

 Tin . 

 Lead . 



German silver 

 Antimony . 

 Bismuth . 



^o 



0.7198 

 C.3760 



0-3435 

 0.2460 

 0.2200 

 0.2041 

 0.1665 

 0.1528 

 0.0836 

 0.0700 

 0.0442 

 0.0177 



'100 



kg X 10'' -{^100 X I05 



0.7226 

 0.3760 

 0.3619 

 O.2S27 

 0.2045 

 0.2540 

 0.1627 

 0.1423 

 0.0764 

 0.0887 

 0.0396 

 0.0164 



45-74 

 24.47 

 22.46 



15-75 

 14.41 

 12.62 



IO-37 

 9-346 



5-141 

 3.766 

 2.199 

 0.920 



33-82 

 17-50 

 17-31 

 13-31 

 10.18 

 11.00 

 6.628 

 6.524 

 3.602 



3-632 

 1.522 

 0-633 



(0, 



*100 _^ £0 

 "10 ^'0 



1574 

 1537 

 1529 



1562 



1527 



I6I7 

 1605 



1635 



1627 



1858 



20U 



1900 



1-358 

 1.398 



1-367 



1.360 



1-315 

 1.428 



1-530 



1-334 

 1-304 

 1-314 

 1.294 



1-372 



C. BeRGET's ExiERlMENTS.t 



The same specimens were used for both experiments. It will be seen that the ratio is nearly constant, but not 



exactly so. 



Substance. 



Copper . 

 Zinc . . 

 Brass . 

 Iron . . 



1.0405 



0-303 

 0.2625 



0.1587 



>t X 10-5 



65-13 



18.00 



15-47 

 9.41 



^lo-s 



1.6 

 1-7 

 1-7 

 1-7 



Substance. 



Tin . . . 

 Lead . . 

 Antimony 

 Mercury . 



0.151 

 0.0810 

 0.042 

 0.0201 



,tXio-= 



8-33 

 5.06 



2-47 

 1.06 



1.8 

 1.6 



1-7 

 1.8 



d. Kohlrausch's Results. 



An interesting confirmation of the relationship of the two conductivities has. been ^ ""^;^':'f ^^Jj f/ ^^^'j 

 rausch, who has shown that tempering steel causes equal proportional changes in the thermal and electrical 

 conductivities of the metal, thus leaving the ratio l/k unchanged by the process.! 



Tempered steel 

 Soft steel 



/= 0.062; >^ = 3-3; l/k = o.o\g 

 " = o.iii; " = 5-5; " =0.020 



In the consideration of this subject it must be borne in mind that closely accurate values of thermal conduc- 

 tivity are very difficult to obtain, and hence fairly large vanations are to be expected. 



* " Wied. Ann." vol. 13, P- So8- 

 t " Compt. Rend." vol. 110, p. 76. 



Smithsonian Tables. 



t /is in c. g. s. units and k in terms of mercury. 



271 



