206 JONES AND SCHUMB. 



The grinding of the stopper was of course made carefully enough so 

 that no grease was needed to make it tight. The material of the cell 

 was sufficiently insoluble so that the specific conductance of the 

 solutions measured did not increase by a significant amomit during the 

 period of the experiment. 



Two cells were used in these determinations, one for thallous 

 chloride, the other for the iodide. In the first cell the electrodes were 

 lightly coated with platmum black; while in the latter cell the elec- 

 trodes were of bright, burnished metal, and placed not over 1 mm. 

 apart. The cell constants of these cells were 0.14658 and 0.02969. 

 The rest of the conductivity apparatus was essentially the same as 

 that already described. Temperature control was very important in 

 the matter of obtaining concordant results; at 0°, especially, it was 

 found necessary that the conductivity cell be deeply immersed in the 

 bath of cracked ice and distilled water. The cell was caused to 

 rotate end-over-end by being attached at right angles to a horizontal 

 shaft. The solid within the cell was thus brought into intimate con- 

 tact with all parts of the solution undergoing saturation. Only a 

 small bubble of air was allowed to remain within the cell on filling. 



Thallous cliloride for these determinations was prepared from pure 

 thallous nitrate, prepared as described on page 201, by precipitating by 

 means of redistilled hydrocliloric acid, followed by repeated washing 

 with conductivity water. The precipitate was not dried, but kept 

 under water in the dark until used. Thallous iodide was prepared 

 from carefully purified thallous nitrate by precipitating with dilute 

 potassium iodide solution. The precipitate is at first flocculent and 

 orange, but on washing and agitation it becomes granular and light 

 yellow. On long exposure to light it takes on a somewhat greenish 

 tinge, but m these experiments the salt was protected against such 

 exposure by a suitable shield above the thermostat and the use of a red 

 heating lamp in the bath. The precipitated thallous iodide was 

 washed with conductivitv water manv times bv decantation, the flask 

 bemg shaken violently so as to break up the larger particles. Like 

 the thallous chloride, it was allowed to soak in conductivitv water in 

 the dark for four months before the rinsing was completed. A small 

 quantity of the moist salt was introduced — more than enough for a 

 number of saturations, however — and rinsed several times with 

 conductivity water; and finally the cell was filled with water, the 

 specific conductance of which was determined at the same time in the 

 "water cell." The rubber caps were affixed to both ends of the cell, 

 and the latter rotated in the constant temperature bath for such 



