Am*De°cU^rarm'}  Colorimetric  Estimation  of  Vanillin.  573 
cessful  because  of  the  filtrate  having  the  peculiar  violet-brown 
coloration;  lastly,  and  with  best  results, -freshly  precipitated  lead 
hydrate  was  agitated  with  the  samples  and  after  standing  a  short 
time  the  mixture  filtered. 
The  lead  hydrate  is  most  quickly  and  conveniently  prepared  by 
dissolving  one  gram  lead  nitrate  in  25-30  cc.  water,  adding  a  few 
drops  of  phenolphthalein  and  then  a  dilute  potassium  hydrate  solu- 
tion until  a  permanent  pink  color  results ;  by  the  cautious  addition 
of  a  dilute  solution  of  lead  nitrate  the  pink  color  is  discharged  and 
then  the  precipitate  washed  two  or  three  times  by  decantation,  and 
lastly  water  added  to  make  100  cc.  of  the  mixture. 
The  determinations  were  made  as  follows :  2  cc.  of  the  samples 
were  diluted  with  about  50  cc.  water,  10  cc.  of  the  lead  hydrate  mix- 
ture added,  and  then  sufficient  water  to  make  100  cc;  after  standing 
a  few  minutes  the  mixture  was  filtered  through  a  dry  filter,  50  cc. 
removed  to  a  flask  of  100  cc.  capacity,  bromine  water  added,  drop 
by  drop,  until  after  agitation  the  bromine  odor  was  permanent,  then 
a  I  per  cent,  ferrous  sulphate  solution  until  the  intensity  of  the 
color  was  reached,  lastly,  water  added  to  make  100  cc.  and,  after 
standing  about  an  hour,  filtered.  One  cc.  extract  by  this  test  will 
form  100  cc.  of  a  more  or  less  colored  solution;  two  standard  solu- 
tions were  made;  one  containing  0-002  and  the  other  0-0005  vanillin 
in  100  cc.  of  the  test ;  the  comparisons  were  made  in  large  test  tubes, 
the  different  height  of  the  columns  being  inversely  proportional  to 
the  amount  of  vanillin  present.  With  colorless  solutions  of  vanillin 
there  is  no  difficulty  in  accurately  determining  the  amount  of  vanillin 
present,  but  in  colored  solutions  there  is  a  yellowish  tinge  to  the  test 
which  renders  accurate  comparison  impossible.  While  this  color 
reaction  is  not  suited  for  the  accurate  estimation  of  vanillin  in 
extracts  and  colored  substitutes,  still  it  allows  a  reliable  compari- 
son of  these  preparations;  this  was  found  to  be  the  case  in  the 
samples  prepared  for  me  by  Mr.  Beringer,  while  the  true  amount  of 
vanillin  would  not  be  given  by  the  test,  still  I  could  invariably  say 
which  contained  more  than  another.  An  important  point  to  be 
noted  in  this  connection  is  that  while  the  odor  of  vanillin  is  destroyed 
in  this  test,  that  of  coumarin  is  not  affected,  so  that  the  test  will 
serve  to  give  an  idea  as  to  the  value  of  a  vanilla  preparation  and 
also  to  detect  small  quantities  of  coumarin  in  presence  of  vanillin. 
Two  of  the  samples  examined  were  from  reliable  manufacturers* 
