^dS™"  fs™  }  Estimation  of  Iodine  in  Kelp  Liquors,  etc.  555 
The  severest  test  that  can  be  applied  to  a  process  of  this  kind,  is 
not  to  use  it  upon  weighed  quantities  of  an  iodide,  but  to  try  it  upon 
liquids  containing,  for  the  quantity  taken,  imponderable  quantities  of 
an  iodide  or  of  an  iodate.  The  experiments  that  have  already  been 
described  on  sea-water,  in  my  paper  "  On  the  Presence  of  Iodate  of 
Calcium  in  Sea-water,"  afford  a  good  illustration.  In  these  experi- 
ments, one  part  of  iodate  of  calcium  in  a  quarter  million  parts  of 
liquid,  sufficed,  in  such  a  quantity  of  the  liquid  as  did  not  contain 
a  ponderable  quantity  of  iodine,  to  give  measurable  iodine  reactions 
in  the  precipitate  thrown  down  by  chloride  of  barium.  And  yet,  if 
we  suppose  the  whole  of  the  iodine  in  sea-water  to  exist  as  iodate  of 
barium,  after  addition  of  a  barium  salt,  the  sea-water  would  contain 
only  about  1  per  cent,  of  a  saturated  solution  of  that  salt.  I  at  first 
supposed  this  very  complete  precipitation  of  iodic  acid  by  barium 
salt  in  sea-water  to  be  owing  to  some  element  contained  in  my  chloride 
of  barium,  that  formed  an  iodate  much  less  soluble  than  any  that  had 
been  described.  I  therefore  took  a  considerable  quantity  of  the  same 
chloride  of  barium  that  had  been  used  in  these  experiments,  and  pre- 
cipitated it  fractionally  by  solution  of  pure  iodate  of  potassium  at 
three  times.  All  the  precipitates  had  sensibly  the  same  degree  of 
solubility  in  water.  Iodate  of  barium  is  remarkable,  however,  in  this  ; 
that  it  dissolves  with  such  extreme  slowness  in  even  boiling  water, 
that  it  is  the  work  of  days  with  the  aid  of  heat,  to  get  a  really  satu- 
rated solution.  The  water  used  in  these  experiments  had  been  care- 
fully distilled  off  permanganate  of  potassium,  since  water  not  perfectly 
free  from  organic  matter  cannot  be  trusted  for  solubility  determina- 
tions of  the  iodates.  The  results  I  obtained  agreed  nearly  enough 
with  the  published  determinations  of  the  solubility  of  iodate  of  barium 
to  convince  me  that  the  complete  precipitation  of  iodic  acid  in  pres- 
ence of  sulphuric  acid  by  a  barium  salt,  is  due  to  surface  attraction, 
between  sulphate  of  barium  formed  in  the  liquid,  and  nascent  iodate 
of  barium.  If  we  consider  bow  many  square  metres  of  surface  a 
single  gramme  of  so  finely  divided  a  precipitate  as  sulphate  of  barium 
usually  is  must  present,  the  property  that  this  salt  is  supposed  to 
enjoy  above  all  others,  of  carrying  down  with  it  more  soluble  salts, 
ceases  to  appear  inexplicable.  Yet  this  power  certainly  is  not  exer- 
cised solely  in  obedience  to  the  degree  of  solubility  of  the  salt  so 
pulled,  by  mechanical  attraction,  as  it  were,  out  of  solution.  A 
colloid  salt  and  a  crystalloid  salt  seem  to  have  no  pulling  power  upon 
