548 Gooch and Read — Determination of Chlorine. 



rotated at about 300 revolutions to the minute, and the current 

 turned on. At the end of the time recorded the anode was 

 removed from the liquid, washed carefully with alcohol and 

 then with ether, dried carefully by slow agitation well above 

 the Bunsen flame, and weighed ; thereafter it was ignited to 

 incipient redness at the top of the Bunsen flame and weighed 

 again, this operation being repeated to insure a constant weight. 

 In the ignition, the dark color of the anode, taken on in elec- 

 trolysis and retained throughout the drying process, largely 

 disappeared. It was determined experimentally that the 

 anode and deposit dried in the manner described incurred no 

 further loss of weight when heated in the air bath at 105°-110°; 

 and a single ignition served to bring the anode and deposit to 

 the second constant weight. 



The liquid remaining after electrolysis was tested with 

 litmus paper and then subjected to the further tests indicated 

 in the accounts of individual experiments. The cathode was 

 examined for deposited silver, and in some cases this was 

 dissolved off the platinum in nitric acid, precipitated as silver 

 chloride, and weighed. The details and results of the experi- 

 ments are summarized in Table I. 



In the experiments of Series A of the table, the apparent 

 error in terms of chlorine, as determined from the weight of 

 the carefully dried anode, varied from —0*0019 grm. to 

 4-0*0022 grm. ; but it is to be noted that the dried anodes lost 

 upon ignition amounts varying from 0'0030 grm. to 0*0070 

 grm., these losses being presumably due to the expulsion of 

 oxygen fixed in the electrolysis. Obviously, the amount of 

 chlorine fixed cannot be calculated from the weights of the 

 dried anodes. If the weights of the ignited anodes be taken as 

 the basis for the calculation of fixed chlorine, the errors of the 

 chlorine determinations vary from —0*0026 grm. to —0*0072 

 grm. 



In every case but one (5), the solution was neutral to litmus 

 paper at the end of the period of electrolysis, which shows 

 that in these determinations no hydrochloric acid remained as 

 such. In each of two cases silver was found upon the 

 platinum cathode to the amounts of 0*0011 grm. and 0*0015 

 grm. respectively, these amounts being determined by dissolv- 

 ing the metallic deposit in nitric acid, and weighing the 

 silver chloride precipitated by hydrochloric acid from the 

 solutions thus obtained. Plainly part of the apparent error in 

 the fixation of chlorine is due in these cases to the transfer of 

 silver from the anode to the cathode. Of the liquids tested, 

 all but one became opalescent upon the addition of silver 

 nitrate, as might be expected if hypochlorous acid or a h} 7 po- 

 chlorite, neither of which would turn blue litmus paper red, 



