366 Messrs. Threlfall and Pollock on the Clark Cell 



Lord Rayleigh's formula is 



E = 1-435 [1-0-00077 (*-15°)J, 



whence it follows that (calling the mean E.M.F. of the 40 

 cells at 15° 0. = 1-43500 volt) at 17° C., the temperature of 

 comparison, the mean E.M.F of the 40 cells is 1-43279 volt ; 

 and hence the E.M.F. of the master cell, as it was 0*00077 

 volt greater than the mean of the 40 cells, = 1*43356 volt. 

 At 21°-5C. the E.M.F. of the master cell (using Lord 

 Rayleigh's coefficient) = 1*42859 volt. 



At the time of the experiments on the large cell the 

 temperature of the laboratory was 21 0, 5 C. and E.M.F. large 

 cell-E.M.F master cell= +0*000405 volt ; therefore E.M.F. 

 of large cell =1*428995 volts at a temperature of 21°'5 C. 

 Thirty-seven minutes after having been put in the bath the 

 E.M.F. of the large cell was steady at 34°*2 0. and E.M.F. 

 large cell -E.M.F. master cell = -0-006885 volt, the master 

 cell being still at 21°*5 C. Hence the fall of E.M.F. of the 

 large cell for a rise of 12°*7 C. was 0*00729 volt, and its 

 actual E.M.F. was 1*4217 volt. 



Hence taking our standard temperature at 21°*5 C, and 

 using Lord Rayleigh's formula with the constant appropriate 

 to 21°*5 C. ; we have 



1*4217 = 1*42900 {1-Kxl2'7}, 



whence K = 0*000402 between 21° and 34° C. say, and for 

 rapid heating. 



There is no reason why this value, deduced from a some- 

 what rapid change of temperature, should agree with the 

 value obtained by Lord Rayleigh in cases where the tempera- 

 ture varied much more slowly and was considerably lower. 

 We have evidence from the history of the 40 cells to the 

 effect that at all events up to 24° C. the differences of E.M.F. 

 are practically the same as they are at 15° C. In the case to 

 which these results have been applied the external resistance 

 through which the cell has to work is composed partly of a 

 copper and partly of a platinoid resistance-coil. The copper 

 coil in this case belongs to the galvanometer and the error 

 thereby introduced, though well within the limits laid down 

 for this particular application, is much greater than it would 

 have been had the galvanometer-winding consisted of German- 

 silver or platinoid wire. There would in this or similar cases 

 have been no objection to the use of such a winding. 



At 15° C, then, to get the most accurate current of 0*001 

 ampere from the large cell we ought to use an external re- 

 sistance of 1430*7 1. ohms, if we ignore the difference between 



