KING. — MEASUREMENT OF THOMSON EFFECT IN COPPER. 367 



of section 9 to 10 when carrying 35 amperes is seen to be 0.0003781, 

 so that the absolute change in resistance on reversal is 



•^^^ X .0003781 = .000000817 ohm. 



85.69 



From column IV, the resistance of the section at 0° is seen to be 

 0.00020304, hence the change of temperature corresponding to the 

 above change of resistance 



.000000817 



= 0°.955 Cu. 



.00020304 X .00422 



The several quantities discussed above, that is, the potentiometer 

 turns representing the whole resistance of the several sections, the turns 

 representing the change of resistance, the actual change of resistance in 

 ohms, the corresponding change of temperature in degrees, are tabulated 

 in columns II., III., VII., VIII., respectively. 



The numbers representing the change of temperature on reversal, for 

 the several sections when plotted, give the curve c d, Fig. 3. 



The h Curve. 



Observations were taken on the five middle sections of the bar, using 

 beating currents of 35, 30, 25, and 18 amperes, in order to obtain short 

 lengths of the temperature distribution curve at several temperatures. 

 Unfortunately the specimen was not long enough to avoid metallic con- 

 duction of heat from the middle section, so that all the heat generated 

 in this section is not lost by radiation and convection. The slight tem- 

 perature gradient at the ends of the middle section, 12 to 13, is shown 

 in Fig. 3. 



Length of section 12tol3 = 1.01 cm. 

 Mean area of cross section = .00720 sq. cm. 

 The thermal conductivity = .870 (1 + .0004 t).* 



With the above data and the temperature gradients at the ends of the 

 section, measm-ed graphically, the amount of heat conducted away from 

 the section for the several values of the heating current is obtained. 



The following table gives the data required for the plotting of the 

 h curve : — 



* From some unpublished experiments. 



