46 JOUENAL OF THE MiTCHELL SoCIETY \_Octoher 



three way stock-cock. W'liile the cell was in tlie thermostat, the? 

 outlet tube was cajoped. The three-way stopcock permitted the 

 removal of either solution, and permitted the separation of the 

 solutions in the limbs of the tube until the measurement was 

 about to be made. The electrodes were of platinum foil about 

 1 cm. square, welded to platinum wire which was fused through 

 a glass tube containing mercury. Frequently during the course 

 of the experiments, the platinum foil and wire were plated with 

 silver from silver nitrate solutions acidified with nitric acid. 

 When the two solutions were at the same level in the two limbs 

 of the U-tube, connection between them was made by opening 

 the stop-cock, and the E. M. F. was determined by the ordinary 

 jDotentiometer method. The galvanometer was sensitive to 

 0.00005 volt even with a large resistance in the circuit. An 

 electrically heated and electrically controlled thermostat was 

 run at 25° constant to 0.01°. The measurements were irregular 

 until the metal coating of the thermostat tank was grounded. 



Measurements of the electromotive force of such combinations 

 were constant within 0.0001 volt for at least 20 minutes after 

 putting the solutions in contact. The table below gives the 

 mean of two values obtained when different solutions and fresh- 

 ly plated electrodes were used. These duplicate measurements 

 differed at most by 0.0003 volt and in the majority of cases by 

 not more than 0.0001 volt. 



These readings show satisfactory agreement among them- 

 selves, for the sum of (1) and (5) should equal (2), and the 

 sum of (3) and (6) should equal (4). The differences are 

 0.0002 and 0.0001 volt respectively. 



1 Gumming finds 59.0 millivolts. 



2 Cumming finds 61.8 millivolts. 



