BISHOP. — THERMAL CONDUCTIVITY OF LEAD. 679 



Procedure. — The temperature of the tanks was adjusted until the 

 difference between them was less than 5° C. The standard Beckman 

 was placed in one tank and Beckman No. 2 in the other. The readings 

 were then taken every two minutes, the galvanometer deflection being 

 taken at the same time. After a series of readings the standard ther- 

 mometer was transferred to the other tank and placed in contact with 

 tl»e sphere as close to Beckman No. 2 as possible. After an interval of 

 sufficient length to insure the two thermometers being at the same tem- 

 perature, their readings were taken, the tank meanwhile being maintained 

 as nearly as possible at its original temperature. If this tank has. not 

 changed in temperature during the comparison of the thermometers, then 

 the difference in the readings of the standard, after correction for calibra- 

 tion, etc., will give the true difference in temperature of the tanks to 

 0.001° Centigrade. If, however, the reading of Beckman No. 2 has 

 changed, it will be necessary to add or subtract this reading to the dif- 

 ference of the readings of the standard. The true temperature of one of 

 the tanks was taken by means of an ordinary thermometer reading 

 directly to tenths of a degree Centigrade. 



Readings on Thermo-Electromotive Force. 



The following set of readings illustrates the method and constancy of 

 the different readings : 



Thermo-Electromotive Force. February 20, 1905. 



All Junctions. Tank No. 1 Higher at 51°.4. 



Readings. 2D. Standard. Beckman. 



79.18 . . 4.289 3.850 



49.31 29.87 4.289 3.851 



79.15 29.84 4.289 3.853 



49.31 29.84 4.289 3.853 

 79.18 29.87 4.289 3.854 

 49.33 29.85 4.291 3.854 

 79.15 29.82 4.292 3.855 



49.32 29.83 4.293 3.855 

 79.18 29.86 4.294 3.858 

 49.30 29.88 4.294 3.857 



29.852 ±.005 4.291 ±0.005 3.854 ± .0004 



