HALL. THERMAL AND ELECTRICAL EFFECTS IN SOFT IRON. 43 



Several temperatures between 0° and 100° were used, the room tem- 

 perature, about 18° , and the boiling-points of ether, bisulphide of car- 

 bon, chloroform, ethylacetate, and benzene, respectively. Using all four 

 of the couples A, B, C, and D in each test, Mr. Churchill found sub- 

 stantially the following mean values of the " thermo-electric height" for 

 the temperatures given alongside : 



Temp. Volt. 



10° C. 1686 x 10- 8 



20° 1734 X 10~ 8 



30° 1782 x 10- 8 



40° 1835 x 10- 8 



50° 1878 X 10~ 8 



60° 1925 x 10- s 



70° 1972 x lO" 8 



80° 2022 x 10- 8 



90° 2073 X 10- 8 



These numbers indicate a very nearly constant rate of increase of the 

 sensitiveness of a copper-germau-silver couple with rise of mean tem- 

 perature on the hydrogen scale. 



As Mr. Churchill left in Cambridge no detailed account of this cali- 

 bration work, it seemed advisable to repeat some of it as a check. 

 Accordingly on November 5, 1904, couples A, B, G, and D were tested 

 again, each with one junction in a mixture of ice and water, and the 

 other junction in steam, each junction being subjected to both conditions 

 in the course of the trial. The result was somewhat alarming, as it 

 indicated for the sensitiveness of the couples a value about 6 per cent less 

 than Mr. Churchill had found; and further study of the matter seemed 

 necessary. 



In the experiment of November 5, as in those of Mr. Churchill, the junc- 

 tions had been exposed naked to the water and to the steam, a test made 

 with one junction in water and one in alcohol, the two being at or very 

 near the same temperature, having appeared to show that no ill effects 

 need be feared from the chemical action or the electrical conductivity of 

 the water. After November 5, however, this practice of direct exposure 

 of the junctions was abandoned, and in all subsequent calibration tests 

 the glass tubes carrying the junctions were placed in glass tubes, about 

 38 cm. long and 0.8 cm. in diameter of bore, containing paraffin e oil ; and 

 these outer tubes were exposed to the' water and the vapors used. The 



