286 



Gooch and Brooks — Rapid method for the 



Evidently the presence of nitrous acid and the treatment by 

 boiling do not interfere seriously with the delicacy of the test. 

 The experiments of the following series were made similarly 

 excepting that potassium iodide was added in varying amount 

 to the solution of the bromide. In these tests a spiral of pla- 

 tinum wire was introduced into the test-tube to prevent too 

 violent ebullition, and at intervals during the boiling the nitrite 

 was added in solution, drop by drop, until the iodine was en- 

 tirely expelled, care being taken to add a little more sulphuric 

 acid toward the end of the separation to ensure completeness 

 of action. The coloiiessness of the liquid in the test-tube after 

 this treatment is a fair indication of the entire expulsion of 

 iodine, but when large amounts of bromide are present free 

 bromine will color the liquid. Salts of iron or other substances 

 which possess color naturally interfere likewise. In such cases, 

 and indeed in all cases, the action of the escaping steam upon 

 red litmus paper is decisive. In the experiments here recorded 

 this test was applied to determine the absence of iodine from 

 the liquid. The residue after the expulsion of the iodine was 

 treated as usual with sodium hypochlorite and shaken with 

 carbon disulphide. 



KI 



KBr 



Total 



Rati 



oof KBr 



Color in 



taken. 



taken. 



Volume. 



to total volume. 



Carbon disulphide. 



0-1000 grm. 



0-0010 grm. 



10 cm/ 



1 : 



10000 



Pronounced. 



o-iooo 



00004 



10 



] : 



25000 



Pronounced. 



o-iooo 



0-0003 



10 



1 : 



33000 



Pronounced. 



o-iooo 



0-0002 



10 



1 : 



50000 



Faint. 



0.1000 



0-0002 



10 



1 : 



50000 



Faint. 



0.1000 



o-oooi 



10 



1: 



100000 



None. 



0-0500 



0-0005 



10 



1 : 



20000 



Pronounced. 



0-0400 



0-0004 



10 



1: 



25000 



Pronounced. 



0-0300 



0-0003 



10 



1: 



33000 



Pronounced. 



0-0200 



0-0002 



10 



1: 



50000 



Faint. 



o-oioo 



0-0001 



10 



1: 



100000 



None. 



o-iooo 



0-00007 



5 I 



1 : 



70000 



Trace. 



0-0070 



0-00007 



5 



1: 



70000 



Trace. 



These results indicate obviously that although the bromme 

 is volatilized to some extent with the iodine it is nevertheless 

 the case that the proportionate amount of bromine removed is 

 dependent upon the absolute amount present, and that enough 

 bromine always remains to permit correct inference as to the 

 quantity originally present, the color test being more or less 

 marked according as much or little bromide was in the solution 

 at the start. The delicacy of the test is a trifle less than that 

 in which the bromide is alone present and not subjected to the 

 treatment which we employ to remove iodine, but the detection 

 of one part of potassium bromide in fifty thousand is certain, 

 and that of one part in seventy thousand possible. It appears, 

 furthermore, that the increase in the amount of iodine 

 present is without effect upon the sensitiveness of the test ; for 

 the bromine in O00007 grm. of potassium bromide was as 



