TOLMAN. — ELECTROMOTIVE FORCE PRODUCED IN SOLUTIONS. 129 



magnitude reached by this residual electromotive force is 0.2 to 0.3 

 millivolts. 



At first sight, it might seem possible to explain this residual poten- 

 tial difference by assuming that the outer electrode was heated more 

 by the friction of the air, during rotation, than the inner one ; since as 

 a matter of fact, a difference in temperature of 1"^ C. would have pro- 

 duced a residual electromotive force of about 0.25 millivolts in the 

 direction actually found.^'* This explanation, however, would not 

 account for the fact that the residual electromotive force tends to in- 

 crease after the machine has been stopped, and the very conditions 

 removed which were supposed to create the temperature difference.^^ 

 This increase in the magnitude of the residual electromotive force 

 after stopping the machine was much easier to follow and the final 

 value reached was much larger, in some earlier experiments where 

 vulcanite tubes were used instead of glass for containing the solution. ^^ 

 This might indicate that the gradual increase of electromotive force 

 after stopping the rotation is due to the gradual emergence into the 

 solution of some constituent which had been forced into the pores of 

 the tube by the centrifugal force or the pressure. Whatever the true 

 explanation, the phenomenon is so complicated that it seemed best 

 not to hazard a guess as to the probable size of the residual electro- 



^^ In order to determine the value of the electromotive forces produced by 

 temperature differences between the electrodes, the thermoelectric power 

 was carefully measured for a number of different circuits of the type, — plati- 

 num ~> salt solution: salt solution — > platinum. The writer hopes to present 

 the results of measurements of this kind in a later paper. 



^' Attempts were made to actually measure the temperature difference 

 between the two electrodes, by placing in the tube an ordinary thermoelectric 

 circuit with iron-nickel junctions at the outer and inner ends. Connection 

 between this thermoelectric circuit and the measuring instruments was made 

 through the mercury contacts already described. It was found, however, 

 that the small electromotive forces arising in these mercury contacts were 

 large enough to obscure those produced by the iron-nickel junctions, and the 

 method was abandoned. There seem to be no practical metallic junctions of 

 higher thermoelectric power than the iron-nickel combination. 



In some earlier experiments very definite tempera,ture effects were produced 

 bj' heat which traveled down the rotator shaft. These effects were eliminated, 

 however, by the copper radiating disk N, Fig. 5, already described, see p. 124- 



^^ In these experiments made with vulcanite tubes, there was also the 

 difference that the rotator was driven in a closed case instead of in the open 

 air. Since the air in the case was considerably heated by the rotation, larger 

 temperature differences might have been expected between the two electrodes- 

 Nevertheless, this would not account at all for the increase in residual elec- 

 tromotive force after the rotator had been brought to rest. 



VOL. XL VI — 9 



