^'^ox1/,^^'}  Separation  of  Iodine,  Bromine  and  Chlorine.  575 
necessary.  Two  equal  volumes  of  the  neutral  solutions,  in  which 
the  halogens  are  to  be  determined,  are  measured  out,  and  one  of 
them  is  accurately  titrated  with  a  1/20  normal  silver  nitrate  solu- 
tion. The  number  of  cc.  required,  a,  and  the  weight  of  the  silver 
precipitate,  b,  are  accurately  determined.  The  second  portion  of  the 
solution  is  treated  with  a  few  grams  of  potassium  bromide,  and  the 
same  volume  of  the  silver  nitrate  solution  as  was  required  to  pre- 
cipitate the  halogens  in  the  first  portion,  is  added.  The  solution  is 
boiled  for  some  time,  diluted  with  water,  and  the  weight,  c,  of  the 
resulting  precipitate,  which  contains  all  the  iodine,  all  the  silver, 
and  some  bromine,  is  noted.  By  means  of  the  three  values,  a,  b  and  c, 
the  quantity  of  each  halogen  present  can  be  readily  calculated. 
If  only  a  small  quantity  of  iodine  and  bromine  is  present  with 
relatively  much  chlorine,  the  method  of  estimation  depends  on  the 
facts  that  silver  iodide  is  insoluble  in  moderately  concentrated  solu- 
tions of  sodium  chloride,  and  that  bromine  and  chlorine  can  be 
separated  by  distillation  with  solutions  of  potassium  permanganate 
and  aluminium  sulphate.  For  the  latter  process,  the  author  employs 
a  distillation  apparatus,  consisting  of  a  retort  and  condenser,  and 
made  of  glass  in  one  piece,  the  last  portions  of  the  bromine  being 
expelled  by  boiling  with  a  little  dilute  sulphuric  acid.  The  bromine 
is  absorbed  in  a  flask  containing  dilute  ammonia,  whereby  it  is  con- 
verted  into  ammonium  bromide  and  probably  partly  into  ammonium 
hypobromite ;  the  whole  of  the  bromine  is,  however,  precipitated  as 
silver  bromide,  on  adding  silver  nitrate  to  the  solution.  This  method 
of  separating  the  iodine  is  only  available  when  (say,  in  a  mixture  of 
sodium  salts)  it  is  present  in  the  proportion  of  not  more  than  I  part 
of  iodide  to  6  or  7  of  bromide  and  1,000  of  chloride,  under  which 
circumstances,  on  the  addition  of  a  little  silver  nitrate,  only  silver 
iodide  is  precipitated,  since  silver  bromide  and  silver  chloride  are 
soluble  in  strong  sodium  chloride  solution.  If,  however,  more  bro- 
mide, or  more  iodide  and  bromide,  are  present  than  is  indicated  by 
the  above-given  ratio,  it  is  best  to  precipitate  and  estimate  the  iodide 
as  palladium  iodide. 
In  making  the  above  separations,  the  author  has  incidentally 
investigated  the  solubility  of  silver  chloride,  bromide,  and  iodide  in 
solutions  of  the  halogen  salts  of  the  alkalies,  more  particularly  in 
sodium  chloride.  Such  solutions  dissolve  4-5  times  as  much  of  the 
halogen  salts  of  silver  at  their  boiling  point,  as  at  the  ordinary  tern- 
