THE 

 LONDON, EDINBURGH, and DUBLIN 



PHILOSOPHICAL MAGAZINE 



AND 



JOURNAL OF SCIENCE. 



[FOURTH SERIES.] 



SEPTEMBER 1863. 



XXIII. On the Conducting-Power of Copper and Iron for Heat at 



o 



different Temperatures. By J. A. Angstrom*. 



§ 1. 1~N a paper sent to the Royal Academy on January 9, 1861, 

 A I described a new method of estimating the conducting- 

 power of metals for heat, by which method that important con- 

 stant is obtained in absolute values (or in values easily reducible 

 into absolute measure), whereas the old methods in use give only 

 relative values. In order still better to establish that the method 

 is practicable, and also to obtain a first estimation of the altera- 

 tion of conducting-power brought about by changes of tempe- 

 rature, I have continued the experiments upon copper and iron 

 which I described in my first paper, having moreover varied 

 them by employing bars of much greater dimensions. The 

 question, whether or not the conducting-power of metals be a 

 function of the temperature, seemed to me to have a not unim- 

 portant bearing, not alone on the theory of heat, but also on a 

 branch of science intimately connected therewith, viz. electricity. 

 This question has been variously handled by former experi- 

 menters. Some of them, as Despretz, suppose that the value 

 of k (the conducting-power) is independent of the temperature, 

 and that the departure from the geometrical law shown in the 

 passage of heat through bars of lead, iron, or other bad con- 

 ductors of heat, is chiefly owing to the thickness of the bar, and 

 the consequent difference of temperature between the different 

 points in one and the same transverse section (a supposition 

 which is in opposition to the foundation of Fourier's theory) ; 

 whilst others, as Langberg, Wiedemann, and Franz, assume that 



* Translated from Poggendorff's Annalen, vol. cxviii. part 3, by Prof. 

 Wanklyn. 



Phil. Mag. S. 4. Vol. 26. No. 174. Sept. 1863. M 



