Force in the Voltaic Arc. 361 



experiments the positive and negative poles were alternately 

 heated. 



Experiment 8. With a somewhat stronger current than in the 

 previous experiments, there was obtained as the mean of ten ob- 

 servations, on 



Heating the positive pole . 40 divisions. 

 Probable error . . . . + 3'1. 



Heating the negative pole . 54 divisions. 

 Probable error .... +3*8. 



The current was also greater in this case when the difference 

 in temperature was diminished. 



Experiment 9. The battery consisted of thirty elements with- 

 out any introduced external resistance. There was obtained as 

 the mean of fifteen observations, 



Without heating .... 46 divisions. 

 Probable error .... ±2'8 



Heating the positive pole . 36*5 divisions. 

 Probable error . . . . +3*1. 



Heating the negative pole . 45*3 divisions. 

 Probable error .... +2*2. 



Experiment 10. The battery consisted of fifty elements, with 

 no other resistance than that afforded by the thick conducting- 

 wires. The current was strong and the deflections great; but 

 the difficulty of obtaining constant deflections increased in the 

 same proportion. There was obtained as the mean of fifteen 

 observations, 



Heating the positive pole . 73 divisions. 

 Probable error .... +7'7. 



Heating the negative pole . 73 divisions. 

 Probable error . . . . +7*1 



Although, as the probable errors adduced indicate, the sources 

 of error are great, the preceding experiments justify the conclu- 

 sion that the current in question does not decrease when the dif- 

 ference in temperature between the poles is less, owing to the 

 heating of the negative pole, which it ought to do if the current 

 were of thermoelectric origin. To this may be added that this 

 diminution of the difference of temperature is very considerable, 

 although it certainly does not entirely disappear. 



When the positive pole is heated, the deflection seems to be 

 somewhat less — though this cannot be considered to be made 



