LEWIS AND WHEELER. — CONDUCTIVITY OF SOLUTIONS. 



421 



trocles (4 sq. cm.) of polished platinum foil were, in the early experi- 

 ments, attached to platinum wires leading through the sides of the tuhes. 

 Later, in order to prevent any pos- 

 sible conduction through the thermo- 

 stat liquid, these wires were carried 

 down inside the tubes from the top 

 and fastened rigidly below by small 

 rods of glass. This arrangement 

 was fairly satisfactory for dilute so- 

 lutions, which attack platinum only 

 slightly and superficially, but with 

 solutions stronger in potassium 

 iodide the platinum suffered seri- 

 ously, the electrode being sometimes 

 nearly eaten away in the course of 

 a few days. For this reason it was 

 desirable to use electrodes which 

 could be removed except during 

 measurements. Electrodes attached 

 to the stoppers, as shown in the 

 figure, proved to be satisfactory in 

 every respect and were used in all 

 the later experiments. Between 

 measurements these stoppers with 

 their electrodes were replaced by 

 plain glass stoppers of similar form. 

 On account of the density of iodine 

 vapor it was found that this could 

 be done without appreciable loss of 

 iodine. The small changes of posi- 

 tion in the electrodes that were 

 possible were shown to have no 

 noticeable effect on the resistance 



of the cell. These movable electrodes had the further advantage that 

 they could be cleaned before each measurement. 



Even with this arrangement, since it was sometimes necessary to leave 

 the electrodes in the solution for an hour or two, it was feared that the 

 platinum might be sufficiently attacked to influence the results. For 

 this reason we used for all concentrated solutions electrodes of platinum- 

 iridium foil and wire, containing 15 per cent iridium. These when first 



Figure 1. 



