302 THE ROYAL SOCIETY OF CANADA 



carbon dioxide, which gives a fairly constant temperature of — 78°C. 

 The chlorine was made by dropping hydrochloric acid on potassium 

 permanganate, was passed through wash-bottles containing water 

 and concentrated sulphuric acid and was dried with phosphorus 

 pentoxide. The mean of a number of measurements gave a value of 

 O.O7X 10~^ for its conductivity, but with the apparatus at our 

 disposal it was impossible to determine the minimum point with any 

 degree of accuracy. 



Tests were made with about 180 inorganic compounds, including 

 various salts of all the common metals, and such substances as water, 

 hydrogen chloride, hydrogen sulphide, tin tetrachloride, bromine, etc. 

 About 80 organic compounds were also tested, including hydrocarbons, 

 alcohols, ketones, aldehydes, esters, acids, amines, nitriles, etc. In 

 no case was evidence of ionization obtained. 



Hydrochloric acid, generated by the action of sulphuric acid on 

 sodium chloride and dried with phosphorus pentoxide, was then 

 bubbled through solutions of pentane, acetic acid, aldehyde, ethyl- 

 amine, ether, alcohol, ethyl acetate and acetone in chlorine. The 

 first four showed no evidence of ionization, but the last four gave 

 conducting solutions. 



Quantitative Measurements 



The conductivity apparatus used was the same as in the previous 

 work but the conductivity vessel was graduated so that the volume 

 of chlorine used in each experiment could be read. Four or five ces. 

 of chlorine were introduced into the cell, the volume read, and the 

 weight calculated from the value for the density. A weighed amount 

 of solute was added from a pycnometer. HCl gas was then slowly 

 passed in from a small gas burette with levelling tube containing 

 mercury, which was protected from the action of the HCl by a layer 

 of concentrated sulphuric acid. The gas was introduced through a 

 capillary tube leading to the bottom of the cell, and after each addition 

 of gas the Dewar fîask containing the cell was lowered sufficiently to 

 remove the capillary tube from the liquid while measurements were 

 being made. Readings of conductivity were taken at regular intervals 

 until the value became constant. It was found that after each addi- 

 tion of HCl it took some time for a constant value to be reached. 

 This time was as much as one or two hours for the first few additions 

 of gas, after which it gradually decreased to ten or fifteen minutes. 

 The change with time was invariably a steady increase except in the 

 case of the toluol solution described later and for a small part of the 

 alcohol curve. 



