444 
IODINE  IN  MINERALS,  PLANTS  AND  ANIMALS. 
reagent  capable  of  indicating  the  presence  of  ~  grm.  of  iodide 
of  potassium  by  means  of  the  blue  color  communicated  to  starch. 
According  to  Grange,  a  few  bubbles  of  hyponitrous  acid,  when 
free  from  nitric  acid,  produce  no  change  in  a  solution  of  pure 
bromide  of  potassium,  while  iodides  are  immediately  decomposed, 
and  the  iodine  separated  may  be  recognized  by  shaking  the  liquid 
with  starch  or  chloroform.  Nitrite  of  potash  and  hydrochloric 
acid  produce  the  same  effect.  The  presence  of  chlorides  does  not 
interfere  with  this  reaction  any  more  than  that  of  bromides. 
Grange'  follows  this  method  in  testing  mineral  and  other  water 
for  iodine,  and  even  for  its  quantitative  estimation.  In  this  case 
the  iodine  is  liberated  by  means  of  hyponitric  acid,  is  separated 
from  the  liquid  by  chloroform,  and  the  colored  solution  treated 
with  a  solution  of  potash  of  known  value,  until  the  color  disap- 
pears, or  else  with  nitrate  of  silver  or  chloride  of  palladium. 
When  chlorine,  bromine  and  iodine  are  to  be  estimated  quan- 
titatively in  a  liquid,  it  is  first  treated  with  hyponitric  acid,  the 
iodine  separated  by  chloroform,  the  bromine,  liberated  by  the 
addition  of  nitric  and  sulphuric  acids,  likewise  dissolved  out  by 
chloroform,  and,  lastly,  the  chlorine  precipitated  by  nitrate  of 
silver.  The  bromine  is  also  precipitated  from  its  chloroform  so- 
lution by  nitrate  of  silver. 
David  Price  applies  the  reaction  between  nitrites  and  iodide  of 
potassium  with  hydrochloric  acid  to  the  detection  of  nitrous  acid 
or  of  iodine.  The  liquid  to  be  tested  for  iodine  is  mixed  with 
starch,  then  with  hydrochloric  acid  and  a  solution  of  nitrite  of 
potash ;  when  much  iodine  is  present  a  dark  blue  color  is  pro- 
duced immediately,  when  it  amounts  to  only  two  or  three  mil- 
lionths  the  color  does  not  appear  until  after  the  lapse  of  some 
seconds. 
With  regard  to  the  sensibility  of  the  various  tests  proposed  for 
iodine,  that  recommended  by  Price  appears  to  be  the  best.  Those 
of  Grange*  and  Marchand  likewise  give  very  trustworthy  results. 
From  a  comparative  study  of  them  I  have  arrived  at  the  follow- 
ing conclusions  : 
1.  The  method  adopted  by  Chatin  and  Gaultier  de  Claubry, 
the  use  of  a  mixture  of  one  part  nitric  acid,  six  parts  sulphuric 
acid  and  starch  powder,  is  only  capable  of  indicating  of 
iodide  of  potassium  by  a  very  faint  violet  color.    Chatin  states 
