BISHOP. — THERMAL CONDUCTIVITY OF LEAD. 689 



varied from 66.21 to 149.97 ; and again, the difference in temperature 

 varied from 3°.034 C. to 7°.844 C. ; consequently we would conclude that 

 the method within these limits is independent of the total energy sup- 

 plied to the inside, and to the difference in temperature. Three different 

 forms of heating coil being used would indicate that the method was 

 independent of the form of the coil. Another point to be noticed is that 

 four preliminary results, taken when the apparatus was set up in the 

 laboratory at Bradley Institute, agree better than 0.5 per cent with the 

 results obtained with the apparatus used in the Ryerson Laboratory. 

 This would imply that the results here obtained are independent of all 

 the electrical apparatus employed. The results for the thermo-electro- 

 motive forces give values and slope of the curve, which are almost identi- 

 cal with the ordinary values given for pure lead and pure nickel. Since 

 the thermo-electromotive force between lead and nickel has been fre- 

 quently determined and with a high degree of accuracy, we may infer 

 that the results here obtained for the thermo-electromotive force contain 

 no constant error, and as shown by the data are probably accurate to 0.1 

 per cent. From which we may infer that the error in measuring the 

 electromotive force when determining the thermal conductivity is also 

 accurate to 0.1 per cent. 



I wish to express my thanks to the Rumford Committee for financial 

 assistance, and to Professor A. A. Michelson and Dr. R. A. Millikan for 

 many courtesies extended to me at the Ryerson Physical Laboratory. 



Physical Laboratory, Bradley Institute, 

 Nov. 4, 1U05. 



VOL. XLI. 44 



