"'/t and Mar— Determination of Chlorine* 297 



hydrochloric acid. At a concentration of LOO cm* the loss is 

 notable. In later experiments, we found that a small part of 

 the precipitate which we weighed in these experiments must. 



in reality have resulted from the action of the Solution Qpon 

 the rubber Btopper and connectors of the apparatus; for the 



distillate from pure water in the same apparatus yielded ji pre- 

 cipitate, probably silver sulphide, which filtered oil* and 



weighed, was found to amount to 0*0003 grm. The figures of 



. therefore, overstate somewhat the volatility of 

 hydrochloric acid under the circumstances, hut the misrepre- 

 sentation is inconsiderable. We fixed upon 300 cm' as a con- 

 venient volume of Liquid to manipulate in future experiments, 

 and one sufficiently dilute to guarantee security against the 

 volatilization of hydrochloric acid when the amount of sul- 

 phuric acid does not exceed 5 cm' of the strong acid, and when 

 the quantity of potassium chloride present does not exceed 

 1 grm. 



The next point to he considered was the proper mode of 

 breaking up hydriodic acid and volatilizing the iodine thus set 

 free, while leaving the hydrochloric acid fixed. 



The use of dihydrogen potassium arseniate, which has been 

 utilized to liberate iodine according to a method recently de- 

 veloped in this Laboratory,* is precluded by the necessarily 

 high degree of dilution of the solution employed. The first 

 trials were made, therefore, with ferric alum to act as the oxi- 

 dizing agent according to the well known reaction of Duflos : 



2H-I + Fe Q (S0 4 ), = 2FeS0 4 + H 2 S0 4 + I-I 



We found by experiment that from a volume of 300 cm 3 

 containing 10 cm 3 of sulphuric acid [1:1], 5 grms. of ferric 

 alum, ami O'OOo grm. of potassic iodide, every trace of iodine 

 had disappeared so completely after five minutes' boiling that 

 nitrous acid and chloroform collected no color in the cooled 

 liquid. It transpired, however, that when the amount of po- 

 dium iodide was increased to 1 grin., iodine was found in 

 considerable amount after boiling for an hour with occasional 

 replacing of^ the water evaporated, so that the volume should 

 not decrease much below 300 cm 3 . The failure of the iron 

 alum to expel the iodine is not attributable to a deficiency in 

 amount ; for the 5 gnn. present were capable of decomposing 

 more than a gram and a half of potassium iodide were the full 

 effect theoretically possible attained. Furthermore, special tests 

 in which the amount of ferric alum was increased developed the 

 fact that the increase was but little helpful. The apparent ex- 

 planation of the phenomenon is that here, as in the liberation 

 of iodine by means of dihydrogen potassium arseniate, there 



*Gooch and Browning, this Journal, vol. xxxix, p. 188. 



