HALL. THERMAL AND ELECTRICAL EFFECTS IN SOFT IRON. 37 



accepted ; though the fact that he does not find a change of slope for the 

 lines of the other substances examined by him may well make us hesitate 

 to reject, as based on experimental errors, his conclusion concerning iron. 



The data of this paper enable us to make a separate calculation of the 

 Thomson effect for the stretch from section c to section m and for the 

 stretch from m to h ; and this calculation is perhaps worth making, though 

 the results will naturally be less reliable than the one result for the 

 whole stretch from c to h. 



Putting, in equation (4), 



S = 0.785 



L = 8.0 



^ = ^ = 0.1543 



l- 2 =k m = 0.1493 

 g<h — g h = A c = 0.3363 

 oi ~ <U = - *m = - 0.00400 

 i — 25.15 



K= 11365 x 10- 9 



y = 0.005 



*. — *,*= 0.21, 



we get, as the first approximation to the Thomson effect heat in the 

 stretch from section c to section m, 



(5') q> = 0.002037 — 0.000235 - 0.000010. 



Correcting for the effect of the higher temperature of the warmer bar in 

 diminishing its thermal conductivity, we get 



(6') q' = 0.002027 - 0.000186 — 0.000010 = 0.001831 ; 



with allowance for lateral leakage we get 



(7') q< = 0.002090. 



To find q", the Thomson effect heat for the stretch m to h, we have 



(7") q" = q -q' = 0.00478 — 0.00209 = 0.00269. 



The value of the Thomson effect coefficient for the stretch c to m 

 would, according to the data pertaining to this stretch alone, be 



