554         Estimation  of  Iodine  in  Kelp  Liquors,  etc.  { k\l™\\  S'"' 
the  effecting  of  improvements  in  the  manufacture  of  iodate  of  potas- 
sium direct  from  the  mother-liquors  of  kelp,  instead  of  from  iodine, 
as  hitherto  practiced),  I  have  found  very  serviceable  for  the  estima- 
tion of  iodine. 
Alkaline  solutions  of  chlorides  and  bromides  are  not  in  the  least 
acted  upon  by  permanganate  solution.  But  neither  chlorides,  bro- 
mides, nor  any  other  salts  that  ordinarily  occur  with  iodides,  inter- 
fere with  the  transformation  of  iodide  into  iodate  by  permanganate. 
Even  organic  matter  does  not  interfere,  provided  the  permanganate 
is  added  in  sufficient  excess.  The  process  I  adopt  consists  simply  in 
adding  excess  of  permanganate  of  potassium  to  the  solution  of  salts 
containing  iodide  until  a  slight  permanent  tint  of  permanganate  col- 
oration remains.  The  solution  is  first  rendered  alkaline,  best  by 
addition  of  caustic  soda,  to  an  extent  adjusted  to  the  proportion  of 
iodide  present ;  but  always  so  far  as  to  preclude  any  possibility  of 
the  liberation  of  iodine.  The  liquid  is  then  filtered,  and  if  it  does 
not  already  contain  a  sulphate,  a  small  proportion  of  a  sulphate  is 
added  to  it.  Solution  of  chloride  of  barium  in  excess — but  not  in 
much  excess — is  then  added,  and  the  precipitate,  after  separation 
from  the  liquid  by  filtration  and  washing,  is  heated  with  solution  of 
sulphate  of  potassium  in  excess.  The  filtered  solution  contains  the 
whole  of  the  iodine  originally  present  in  the  portion  taken  for  analy- 
sis, as  iodate  of  potassium.  The  quantity  of  iodic  acid  may  be  esti- 
mated volumetrically  by  any  of  the  usual  processes ;  or  the  mixture 
of  iodate  and  sulphate  may  be  ignited  at  a  low  red  heat,  and  the 
iodide  of  potassium  remaining  be  estimated  either  volumetrically  or 
gravimetrically. 
In  this  process,  the  transformation  of  iodide  into  iodate  by  per- 
manganate in  alkaline  solution  is  complete.  The  precipitation  of  the 
iodic  acid  by  a  barium  salt  in  presence  of  a  sulphate  is  complete. 
The  decomposition  of  iodate  of  barium  by  heating  with  solution  of 
sulphate  of  potassium  in  excess,  is  complete.  By  the  word  "  com- 
plete," I  mean  within  appreciable  limits.  That  is  to  say,  I  have,  in 
the  couise  of  many  experiments  specially  devoted  to  that  end,  been 
unable  with  certainty  to  detect  even  a  trace  of  iodine  either  in  a 
liquid  in  which  an  iodide  had  been  transformed  into  iodate  as  de- 
scribed, and  precipitated,  in  presence  of  a  sulphate,  by  chloride  of 
barium ;  or  in  the  barium  precipitate,  after  heating  with  sulphate  of 
potassium  in  excess,  filtering,  and  duly  washing. 
