376 The Clark Cell wlien 'producing a Current. 



My results when the current was maintained indicate that 

 the potential-difference at the electrodes became less. In my 

 earliest experiments, when two Leclanche cells were used in 

 the potentiometer circuit, I found an effect similar to that 

 observed by Professor Threlfall. This is stated in my paper, 

 § 5 and § 7, and as such a result was not expected and did 

 not for many reasons seem to be connected with the Clark 

 cell, I placed large Clark cells in the place of the Leclanches. 



The considerations which led me to make this change may 

 be shortly summarized. Firstly, the comparison of the size 

 of the plates in the two cells is greatly in favour of the large 

 Clark, and also the soluble depolarizer in that cell would cause 

 more efficient depolarization. Secondly, I have frequently 

 observed when using the potentiometer that the same value 

 may be obtained for the potential-difference of a short- 

 circuited Clark before and after an interval during which the 

 Clark has been on open circuit, provided no large plug change 

 has been made during the interval. On the other hand, a 

 large plug change, such as testing the E.M.F. of the cell with 

 its poles open will lead to a new value when its poles are 

 again closed. Thirdlv, when determining the absolute value 

 of the E.M.F. of the Clark cell with Mr. Glazebrook (Phil. 

 Trans. 1892), we thought it safer to avoid large plug altera- 

 tions, and so abandoned this process to adopt one in which a 

 mercurial rheostat in the main circuit could be varied so as to 

 maintain the electrodes of the standard resistance at a constant 

 difference of potential. With this alteration my experiments 

 became regular ; seven sets were obtained (one lasted 16 

 hours), and I was satisfied that the original irregularity was 

 due to the Leclanches. The average of four sets is given in 

 my paper. If I w r ere to repeat these measurements 1 should 

 now T use one large accumulator and a wire rheostat for that 

 part of the potentiometer where alterations of resistance would 

 have to be made, and thus avoid all rapid plug chances. This 

 arrangement would, I am sure, give better results. 



My object in referring to Mr. Fitzpatrick's paper in the 

 easily accessible B.A. Reports w 7 as to avoid a lengthy descrip- 

 tion of an apparatus which is in continuous use in the 

 Cavendish Laboratory. In his paper Mr. Fitzpatrick has 

 made the proper references to Macgregor's and Professor 

 Kohlrausch's well-known work. 



I am. Gentlemen, 



Yours respectfully, 



S. Skinner. 

 Cambridge, March 9, 1895. 



