462 THE LIFE OF JAMES 7). FORBES. [CHAP. 



conductivity is the same as that for electric conductivity; 

 and one grand result of his new inquiries was to show 

 that, in iron at least, increase of temperature diminishes 

 the thermal conductivity as it had long been known 

 to diminish the electric conductivity ; and, curiously 

 enough, in something approaching to the same propor- 

 tion. [M. Langberg, of Christiania, writing to Forbes in 

 November 1853, says: 'Some days since I got the Eeport 

 'of the British Association for 1852, from which I see 

 ' that you have experimentally proved the important fact 

 'that the conductivity of iron for heat diminishes as the 

 'temperature increases. I have felt much interest in 

 ' reading this statement, as I think I have myself shown 

 ' (Pogg. Annalen, v. 66, p. 1, 1845), that the conductivity 

 ' of tin, lead, and steel, is a function of the temperature ; 

 1 but the nature of my experiments did not allow me to 

 'find if it increased or diminished with the tempo - 

 'rature.'] The method employed by Forbes, which was 

 entirely devised by himself, was a singular improvement 

 upon that adopted by Biot and others, who, having 

 assumed the conductivity to be unaltered by temperature, 

 and having adopted for all temperatures an empirical 

 law of cooling by radiation and convection, deduced by 

 Fourier's methods the corresponding stationary distribu- 

 tion of temperature ; and then, by the method of least 

 squares, endeavoured to make their observations agree 

 with it. Forbes deduced, by separate experiments, the 

 rate of cooling of his bar for all temperatures afterwards 

 to be observed. He also carefully measured the distri- 

 bution of temperatures in the stationary experiment, 

 and used Fourier's method only for elementary portions 

 of the bar. His results are of great value, and accord 

 well with those since obtained by Angstrom by a totally 

 different, but also very ingenious, process. 



Another result of the same investigations, and one 

 of great interest and importance in modern science, is 

 his determination the earliest of much real value of 

 the absolute conductivity of a metal ; i.e., how much 

 heat passes per second per unit of surface through an 



