422 PROCEEDINGS OF THE AMERICAN ACADEMY. 



used were slightly etched on the surface, hut afterwards they appeared to 

 he unattacked even after long standing in concentrated solutions. 



The cells were immersed nearly to the top in the thermostat, which 

 was of the ordinary type and so arranged that it could be readily set 

 at any one of three temperatures, 120°, 140°, and 160°. At first 

 cotton-seed oil was used for the bath, but was later discarded, partly on 

 account of the vapors which, rising from it and condensing on the sides of 

 the cell, endangered the purity of the contents, and partly on account of 

 the gradual hardening which it underwent. After some experimenting a 

 suitable substance was found in a commercial substitute for lard known 

 as cottolene. 



The method of carrying out a series of experiments was as follows : 

 Enough iodine (60-80 gm.) was weighed out to stand when melted 2 or 

 3 cm. above the electrodes. The glass stoppers were inserted at once to 

 prevent loss of iodine, the cell placed in the thermostat, and its resistance 

 determined. Then the stoppers were removed, after gently warming 

 them with a flame to melt down the small quantity of iodine which sub- 

 limed into the space between the stopper and the tube. A known weight 

 of potassium iodide, pulverized and pressed into a small pellet, was intro- 

 duced, and the liquid was stirred by a simple apparatus which drew it 

 back and forth from one arm of the cell to the other. The stoppers were 

 replaced and the resistance again determined. In the same way more 

 iodide was added and the process repeated. At each concentration the 

 resistance could be determined for a number of temperatures. 



The measurements of resistance were made by means of a Wheat- 

 stone's bridge, both with direct current and galvanometer, and with 

 alternating current and telephone. The direct current was very con- 

 venient and accurate, and with dilute solutions gave excellent results. 

 The two electrodes showed no potential difference either before or during 

 the passage of the current. * 



In order to calculate the specific resistance of the solution from the 

 observed resistance, the constant of each cell was obtained by means of a 

 standard aqueous solution of potassium chloride or potassium iodide. 



The only chemicals used in the research were potassium iodide and 

 iodine. The former was purified by several recrystallizatious. Two 

 kinds of iodine were used ; one was the pure resublimed iodine prepared 



* This lack of polarization is no evidence that the conductivity is metallic in 

 character. The same phenomenon is observed with an aqueous solution of potas- 

 sium iodide and iodine. 



