246  x  The  Estimation  of  Free  Acid.        { Am  jS^um™1' 
solution,  50  Cc.  water,  and  2  Cc.  decinormal  sodium  hydroxide 
solution.  To  the  contents  of  one,  five  drops  of  a  ten  per  cent,  solution 
of  cupric  chloride  are  added.  After  one-half  hour's  standing,  9  Cc. 
of  a  normal  solution  of  sodium  thiosulphate  are  added  to  the  contents 
of  each  flask.  Both  liquids  are  immediately  colored  deep  violet.  The 
copper- free  solution  decolorizes  very  slowly;  after  five  minutes  the 
color  is  Madeira  brown,  and  after  another  five  minutes  or  more  the 
liquid  becomes  almost  colorless  and  deposits  a  precipitate  of  ferric 
hydroxide. 
The  solution  containing  the  cupric  chloride  decolorizes  almost 
instantly,  and  becomes  turbid  in  a  very  brief  space  of  time. 
When  the  decinormal  sodium  hydroxide  was  replaced  by  hydro- 
chloric acid  a  clear  acid  solution  was  obtained. 
In  this  liquid  the  free  acid  may  be  approximately  determined  * 
by  titration,  using  methylorange  as  an  indicator.  The  estimation 
is  not  exact,  as  the  deposition  of  ferro-ferric  hydroxide  begins  early 
in  the  operation,  covering  the  change  of  color  and  deflecting  a  por- 
tion of  the  base  from  the  titration.  This  fault  may  be  largely  over- 
come by  adding  a  starch  solution  as  protecting  colloid.  Other  col- 
loids may  give  better  results  or  may  react  in  such  a  manner  as  not 
to  be  serviceable  for  this  purpose. 
I  have  obtained  good  results  with  Kahlbaum's  soluble  starch  and 
with  soluble  starch  prepared  after  the  method  of  Fernbach  and 
Wolf.1 
The  following  reagents  are  recommended  for  the  estimation : 
Normal  sodium  thiosulphate  solution  containing  24.8  grammes 
sodium  thiosulphate  in  100  Cc. 
Cupric  chloride  starch.  Mix  one  gramme  cupric  chloride  with 
49  grammes  of  soluble  starch  previously  dried  at  ioo°  C. 
The  titration  is  effected  as  follows : 
To  the  cooled  solution  of  0.5  gramme  cupric  chloride  starch 
in  50  Cc.  of  water,  contained  in  an  Erlenmeyer  flask  of  100  Cc. 
capacity,  add  2  Cc.  of  the  ferric  chloride  solution  to  be  tested.  Nor- 
mal thiosulphate  solution  is  now  added,  5.5  Cc.  for  U.  S.  P.  ferric 
chloride  solution  and  9  Cc.  for  that  of  the  P.  N.  IV. 
The  decolorized  liquid  is  colored  fairly  strong  with  methyl- 
orange  solution.  If  the  reaction  of  the  mixture  is  acid,  it  is  titrated 
drop  by  drop  with  decinormal  sodium  thiosulphate. 
1  Zeitschrift  f.  analytische  Chemie,  1914,  p.  708. 
