198 



(j-ooch and Browning — Method for the 



In Series K are detailed experiments which follow the line 

 marked ont for the separation of iodine from chlorine and bro- 

 mine in the haloid salts, excepting that the iodide was entirely 

 omitted for the purpose of discovering whether hydrochloric 

 and hydrobromic acids possess reducing action upon arsenic 

 acid under the conditions. The evidence is plain that no arseni- 

 ous oxide is formed by the action of 0*5 grin, of sodium 

 chloride upon 2 grm. of the arseniate under the circumstances 

 as given. 



Hydrobromic acid, on the contrary, is slightly decomposed 

 with the evolution of bromine enough to give visible color to the 

 concentrated liquid, but the amount of bromine lost, as indicated 

 by the arsenious acid produced in its evolution, is only 0*0003 

 grm. for 0*5 grm. of potassium bromide, and 0*0001 grm. for 

 0*1 grm. of the bromide. Further experiments indicated, 

 however, that concentration cannot go on to a volume less 

 than 35 cm 3 without causing serious loss when the maximum 

 amount of bromide is present. 



Series K. 













Iodine cor- 



Chlorine 



Bromine 



H 2 S0 4 

 [1:1]. 









Final 



resoondi'g 



corres- 



corres- 



H 2 KAs0 4 . 



NaCl. 



KBr. 



volume. 



to As 2 3 



ponding 



ponding 











reduced, 

 grm. 



(average), 

 grm. 



(average )u 













grm. 



20 5m 3 



2 grm. 



0*5 grm. 







35 cm 3 



0-0000 ) 







20 



2 



• 0-5 



_ _ 



35 



0-0000 \ 



o-oooo 



____ 



20 



2 



0-5 





35 



o-oooo ) 







20 



2 





1 grm. 



35 



0-0003 ) 







20 



2 





o-i 



35 



o-oooi J. 







o-oooi 



20 



2 





o-i 



35 



o-oooi ) 







20 



2 





05 



35 



00005 ) 







20 



2 





0-5 



35 



0-0005 [ 



_ 



0-0003 



20 



2 





0-5 * 



35 



00005 ) 







In Series L are given the results of twenty- six determinations 

 of iodine by the method outlined. 



Upon inspection of these results it appears that the method 

 is good and reliable under all the conditions tested. 



When neither chloride nor bromide is present, the iodine is 

 determinable with a mean error of 0*0002 grm. 



When sodium chloride accompanies the iodide, the results 

 show a loss of arsenious oxide and a consequent apparent defi- 

 ciency of iodine — as we should expect in accordance with the 

 indications of Series I. This deficiency proves to be propor- 

 tional to the amount of iodide broken up, or the arsenious oxide 

 thus produced. For 0*56 grm., approximately, of potassium 

 iodide and 0*5 grm. of sodium chloride the deficiency meas- 

 ured in iodine amounted to 0*0011 grm. When the potassium 

 iodide is decreased ten-fold (or more) the deficiency falls to 



