ON THE HALOGEN HYDRIDES AS CONDUCTING SOLVENTS. 145 



manner previously described and 10 cub. centims. of the solution were taken out and 

 analysed. This analysis gave the concentration of the solution before electrolysis. 

 A silver voltameter and a milliampermeter having been placed in the circuit, the 

 electrolysis was started by connecting the electrodes to the terminals of a battery, 

 the voltage of which could be varied within wide limits, and the voltage was adjusted 

 so as to give a current of not more than 12 or 14 milliamperes. If a larger current 

 than this was employed, the heating effect was found to cause mixing of the liquid 

 by convection. After the electrolysis had been continued for about two hours it was 

 stopped by removing the cathode from the voltameter, and the liquid in the cell was 

 separated into two portions by inserting the stop-cock ,s into its socket. An absorber 

 was then attached to the transferrer T' and the column of liquid contained within 

 the dotted lines was blown into the absorber, which was then removed. The third 

 absorber was next attached to the transferrer T and the solution contained in the 

 apparatus forced into it by blowing through w, the tube r acting as a syphon to 

 remove the liquid contained in P, and finally the third absorber was removed and 

 weighed. Three solutions were thus obtained for analysis, namely: (1) the 10 cub. 

 centims. which had been removed before electrolysis and gave the original concen- 

 tration ; (2) the small quantity taken out in the second absorber (which should be of 

 the same concentration as (1)); and (3) the solution surrounding the cathode, which 

 gave the change of concentration brought about by the electrolysis. 



In order that an experiment should be successful it was necessary that the foregoing 

 procedure should be strictly followed. Identical values for the original concentration 

 and that of the middle portion have never been obtained unless the whole solution 

 was first placed in the apparatus and all three portions were taken from it. 



At least half-a-dozen other methods have been tried without success. Fortunately 

 it is easy at the close of an experiment to see if any mixing has taken place, from the 

 fact that bromine is liberated at the anode, where it forms a deep red solution in 

 the hydrogen bromide. If the experiment has been successful, this solution remains 

 as a very clearly defined layer surrounding the anode, and the coloration does not 

 extend more than about 1'5 centims. up the tube. On the other hand, if mixing has 

 taken place, as may happen either if the current is too large or if the temperature of 

 the bath is allowed to vary, the bromine is distributed throughout the solution and 

 no clearly defined layer is seen at the anode. 



Experimental Results. 



The results of the experiments are contained in the following table, in which the 

 concentration of the various solutions is expressed under N, which gives the number 

 of gramme equivalents of dissolved substance per litre of solution, the percentage also 

 being given in the 3rd column. The 4th column contains the weight of silver 

 deposited on the cathode of the silver voltameter, the 5th column gives the weight of 

 substance transported, the 6th column the cation transport numbers. 



VOL. cov. A. u 



