HALL. THERMAL AND ELECTRICAL EFFECTS IN SOFT IRON. 39 



To establish a uniform gradient the intermediate values should have been 



at 41.3° 975 X 10~ 8 



« 54.5° 029 X 10-« 



With these data, and with the value of v found in this paper for iron, it 

 is possible to go through the form, at least, of calculating the value of v 

 for copper. The process is as follows : 



E. m. f. of copper-iron couple with junctions at 26.6° C. and 71.1° C, 

 respectively, is 



E = 1028 + 8 '° x iQ-8 x (7i.i _ 2G.6) volt = 422 X 10~ 6 volt. 



Heat, calories, which would be absorbed by this net e. m. f. in one second 

 by 10 amperes is 



q n = 422 X 10-° X 10 X 0.2387 = 0.00101. 



Heat, calories, which would be absorbed at the hot junction in one second 

 by 10 amperes is 



q h = 870 x 10- 8 X (273 + 71.1) X 10 X 0.2387 = 0.00715. 



Heat which would be given out at the cold junction by the same current 

 in the same time is 



q c = 1028 X 10- 8 x (273 + 26.6) X 10 X 0.2387 = 0.00735. 



Hence the heat which would be absorbed in the Thomson effect in both 

 metals by the same current in the same time is 



q T = 0.00101 - (0.00715 — 0.00735) = 0.00121. 



Letting v c and v t stand for the Thomson coefficients in copper and iron 

 respectively, we have 



Vc _ v . = q T + (71.1 _ 26.6) ^273 + 26 - G + 71A \ = 844 x xq-w 



Taking v, = — 757 X 10 -10 , the mean value found in this paper, 

 we get 



v c = 844 x 10- 10 - 757 X 10- 10 = 87 X 10" 10 . 



King, in the paper already referred to, gave evidence from which he 

 concluded that the Thomson effect o- for copper diminishes with rise of 

 temperature. The v would of course diminish still more rapidly. He 



