268 



PROCEEDINGS OF THE AMERICAN ACADEMY 



End no. 1 heated, 

 June 11, 1890 



End no. 2 heated, 

 June 11, 1890 



Mean 



Impure Cast Nickel. 



Conductivity (c. g. 8.) 



.108 / 1Q65 at 



.105 S 

 .1030 > 1062 

 .1095 ) 



. . . .106 



Temperature 



127° C. ) 



115 ) 

 128 I 



116 S 



116 



117 



116 



The avowedly rough and hasty character of the experiments 

 being considered, the agreement of the results obtained in the case 

 of gun-iron and in the case of nickel is entirely satisfactory. In 

 the case of the other bar, the results do not agree among themselves 

 so well as one could wish, although, in view of the acknowledged 

 difficulty which attends experiments upon thermal conductivity, 

 they should not be considered surprisingly discordant. I can offer 

 no explanation of the fact that both the largest and the smallest 

 calculated values of the conductivity were obtained from this bar. 

 The mean value, it will be observed, is not very far from the mean 

 obtained with the gun-iron bar. The mean value for each of the 

 iron bars is, I believe, smaller than any other investigator has 

 recorded for iron, though not smaller than has been found for 

 Bessemer steel. 



The following values are taken from Landolt and Bornstein's 

 Physikalisch-Chemische Tabellen : — 



According to Kirchhoff and Hansemann ( Wied. Annalen, Vol. 

 IX. p. 5), the conductivity of iron (in c. g. s. units) is expressed 

 by the formula .1694 — .00034 (t — 15). 



This gives for 15° C. the value .1694, 

 and for 100° C. the value .1405. 



The iron used by Lorenz had the density 7.828 {Wied. Annalen, 

 Vol. XIII. p. 438), and was therefore not cast-iron. That used by 



