290 Messrs. Wright and Thompson on the Determination of 



In order to determine these limiting current-densities, 

 experiments were made in the following way. The diffusion- 

 cell, or pair of opposed Daniell or other cells to be examined, 

 was set up, as was also a normal Daniell cell for compa- 

 rison, and the currents from each alternately sent through 

 a large external resistance (using appropriate shunts), series 

 of readings being taken in each case. The resistance was 

 now altered, and two fresh series of readings taken ; after 

 which it was again altered, and two more series taken, and 

 so on : usually the whole set of series of readings was 

 repeated in reverse order. The average scale-readings for 

 a series of given total resistances in circuit were thus known 

 both for the Daniell cell and the other one compared with it. 

 With the former, the product of average scale-reading and 

 total resistance was sensibly constant for current-densities 

 that caused an appreciable falling-off in the product similarly 

 obtained with diffusion-cells or with two opposed Daniell 

 cells, &c, showing that whilst " polarization " did not occur 

 to a measurable extent in the first case, it did in the second. 

 The following illustration will make the point clear : two 

 series of readings with varying resistances were made in 

 reverse order with a normal Daniell cell, and with two 

 opposed cells jointly giving an effective E.M.F. of a little 

 more than one-hundredth of that of the Daniell cell; for the 

 former the yoW shunt was used, and for the latter the ^ 

 shunt. 



Single Daniell Cell. 



Total resist- 

 ance in circuit, 

 in ohms. 



Scale-readings. 



Product of mean 



scale-reading 

 into resistance. 



First 

 series. 



Eeversed 

 series. 



Mean. 



200,050 

 150,050 

 100,050 



63-7 



84-8 



127-5 



63-9 



85-3 



127-6 



63-8 



85-05 

 127-55 



12-763 Xl0 6 

 12-762 xlO 6 

 12-761 x 10 6 



Here the product of scale-reading into resistance being 

 constant within the limits of observational error, it is evident 

 that the E.M.F. of the cell was constant within about 20*00 

 part of its value, notwithstanding that the current rose from 



I'll fK k • v * I'll 



200050 am P ere ( 5 ' 5 microam peres) to jqqqjq ampere 



(11*1 microamperes), representing a rise in density of from 2 

 to 4 microamperes per square centimetre, the plates exposing 

 a surface of about 3 square centimetres each. Each scale- 



1*115 



division here represented iqaa x 1 9.7^0 — :ttt 6 ampere,or87'37 



elassoamperes. 



