312  Volumetric  Determination  of  Mercury.  {^^'/X^iu'^^' 
Erlenmeyer  flask,  add  50  c.c.  of  n/io  iodine  solution.  Mix  by 
rotating  until  the  salt  is  thoroughly  moistened,  then  add  a  solution 
of  2  Gm.  of  potassium  iodide  in  10  c.c.  of  water  and  at  once  shake 
the  stoppered  flask  vigorously  until  solution  is  complete.  Titrate 
the  excess  of  iodine  with  n/io  sodium  thiosulphate,  using  starch 
solution  as  indicator,  adding  the  latter  when  the  liquid  is  nearly 
decolorized.  Each  c.c.  of  n/io  iodine  solution  consumed  corre- 
sponds to  0.02355  Gm.  of  mercurous  chloride. 
A  sample  of  calomel,  practically  chemically  pure,  assayed  by  this 
method  99.5  to  100.0  per  cent.,  the  average  of  6  determinations  being 
99.7  per  cent.,  with  some  variation  in  the  working  details.  Prac- 
tically the  same  figures  were  obtained  with  yellow  mercurous  iodide 
containing  slight  traces  of  impurities. 
The  criticism  of  the  Rupp  method,  that  reduction  of  the  mercury 
is  not  complete  within  a  reasonable  time  without  the  use  of  heat 
and  that,  when  heat  is  employed,  the  mercury  is  not  easily  dissolved 
afterwards,  is  well  founded,  if  Rupp's  directions  in  outhning  the 
method  (Berichte  d.  d.  ch.  Ges.,  1906,  v.  39,  p.  3702),  be  taken 
literally  as  regards  the  quantity  of  alkali  to  be  used.  While  his 
directions  lead  to  the  inference  that  only  enough  is  required  to  com- 
bine with  the  acids  formed  by  the  reaction,  his  figures  in  the  same 
paper  show  that  he  used  a  decided  excess,  not  less  than  10  c.c.  of 
normal  solution  for  0.2  Gm.  of  mercuric  chloride.  The  divergent 
opinions  regarding  this  method  by  the  members  of  the  committee 
are  doubtless  due  chiefly  to  differences  in  the  quantities  of  alkali 
used.  When  the  above  mentioned  proportions  are  taken,  the  reac- 
tion is  almost  instantaneous,  unless  the  solution  is  excessively 
diluted,  and  may  safely  be  regarded  as  complete  within  5  minutes, 
without  heating.  It  will  do  no  harm  to  use  a  still  larger  excess  of 
alkali  or  to  let  the  mixture  stand  longer.  Shaken  in  a  stoppered 
flask  with  the  iodine  solution,  the  precipitate  is  then  dissolved  very 
readily.  A  large  excess  of  acetic  acid  is  to  be  avoided,  as  it  tends 
to  make  the  result  too  low,  which  is  also  the  experience  of  Mr.  L.  D. 
Havenhill,  of  the  committee.  The  best  results  were  obtained  by 
carrying  out  the  details  as  follows :  Dissolve  about  0.5  Gm.  of 
powdered  mercuric  chloride,  accurately  weighed,  contained  in  a 
glass-stoppered  300  c.c.  Erlenmeyer  flask,  in  a  solution  of  2  Gm.  of 
potassium  iodide  in  10  c.c.  of  water.  Add  25  c.c.  of  normal  caustic 
alkali  solution  and  6  c.c.  of  40  per  cent,  formaldehyde  solution. 
Mix  by  swirling  the  flask  occasionally  during  10  minutes,  then  acidu- 
