550  Benzoic  vs.  Cinnamic  Acid.  1^X^19™" 
the  fluid  with  ammonia  water,  and  adding  a  drop  or  two  of  ammo- 
nium sulphide  solution,  when  a  red  color  appears  if  benzoic  acid 
was  present. 
Neither  of  these  tests  will  distinguish  between  benzoic  and  cin- 
namic acids.  In  fact,  cinnamic  acid  responds  to  both  tests  almost 
as  well  as  benzoic. 
With  ferric  chloride,  in  moderately  strong  and  colorless  solutions, 
cinnamic  acid  gives  a  precipitate  that  is  decidedly  yellowish,  and  is 
distinguishable  from  benzoic.  But  in  weak  solutions,  when  the  pre. 
cipitate  is  slight,  yet  marked,  the  precipitates  cannot  be  distinguished 
by  color,  and  when  the  test  is  applied  to  a  deeply  colored  solution 
as  is  obtained  from  sweet  pickles,  it  is  obvious  that  a  test  depending 
upon  a  tint  cannot  be  trusted. 
Some  oil  of  cinnamon  (both  cassia  and  Ceylon  cinnaman  oils  were 
used  in  separate  tests)  was  shaken  with  warm  water  and  the  liquid 
filtered.  A  drop  of  ferric  chloride  added  to  a  test-tube  full  of  the 
filtrate  produced  a  precipitate  that  could  not  be  told,  by  color  or 
physical  appearance,  from  that  produced  in  a  weak  solution  of 
sodium  benzoate  by  ferric  chloride.  Yet  the  conditions  of  the  test 
were  most  favorable  for  comparing  results.  Some  cinnamon  water 
was  distilled  with  steam,  the  distillate  neutralized  with  ammonia, 
evaporated  to  a  small  volume,  rendered  acid,  shaken  out  with  ether, 
the  ethereal  solution  shaken  with  water,  then  with  weak  ammonia 
solution,  and  the  excess  of  ammonia  driven  off  by  evaporation  on  a 
water  bath.  This  solution  gave  a  precipitate  with  ferric  chloride 
that  might  fairly  be  taken  for  benzoic  acid ;  and  this  process  was 
the  one  used  by  the  Government  chemist,  and  upon  which  the 
indictment  was  based. 
Some  cinnamic  acid  was  prepared  by  oxidizing  oil  of  cassia  with 
hydrogen  peroxide,  and  the  purified  acid  (having  a  melting  point  of 
1 3  3°  C),  reacted  to  both  the  iron  and  the  ammonium  sulphide  tests 
in  a  way  similar  to  benzoic  acid. 
It  is  obvious  that  these  tests  cannot  be  relied  upon  to  detect 
benzoic  acid  in  a  product  that  is  flavored  with  cinnamon. 
The  best  test  found  for  distinguishing  between  the  two  acids  is 
the  manganous  test.  Manganous  salts  give  a  white  precipitate  with 
cinnamates,  which  gradually  becomes  crystalline.  No  precipitate  is 
found  with  benzoates,  even  in  moderately  strong  solutions. 
The  precipitate  forms  slowly  with  cinnamates,  but  is  quite  delicate 
