CONTRIBUTIONS FROM THE JEFFERSON PHYSICAL LABORATORY, 

 HARVARD UNIVERSITY. 



MEASUREMENT OF VARIOUS THERMAL AND 



ELECTRICAL EFFECTS, ESPECIALLY THE 



THOMSON EFFECT, IN SOFT IRON. 



By Edwin H. Hall, Messrs. Churchill, Campbell, and Sekviss.* 



Presented December 14, 1904. Received March 9, 1905. 



Part I: General Account of Methods, Apparatus, and 



Results. 



Let the end E of a bar of metal be maintained at some fixed tempera- 

 ture, T, and let the end E of the same bar be maintained at a different 

 fixed temperature, V. If an electric current of given strength is then 

 driven through the bar, the amount of heat generated by this current 

 will depend on the direction of the current, being either greater or less 

 when the current flows from E to E' than when if flows in the opposite 

 direction. This directional relation between an electric current and the 

 heat generated by it is the simplest and most direct aspect of the 

 " Thomson effect." 



Let us suppose that we have initially, before the electric current is put 

 on, a uniform ascending gradient of temperature from the end Eto the end 

 E', indicated by the straight line S in Figure 1. Let J in the same figure 

 represent the line of temperature gradieut which would be established in 

 the bar by a permauent electric current of strength i, if there were no 

 Thomson effect. Let the line A represent the temperature gradient 

 which would actually be maintained in the bar by the permanent current 



* Most of the preliminary work of this investigation, making the main thermo- 

 electric couples, testing them, varnishing the iron bars and fixing the thermo-elec- 

 tric couples upon them, assembling the apparatus as a whole and getting it into good 

 working order, was done by Mr. Churchill, who gave much time to the undertaking 

 during the academic years 1902-3 and 1903-4. Unfortunately Mr. Churchill, leaving 

 Cambridge early in September, 1904, was unable to take much part in the main 

 observations, which were made during the fall of that year and the early part 

 of 1905, largely by Mr. L. L. Campbell and Mr. S. B. Serviss of the Harvard 

 Graduate School. 



