20 
A  m  m oniw n  B ens oat e. 
(Aru.  Jour.  Pharui 
)    January,  1910. 
be  obtained  by  varying  the  rate  of  heating  or  by  holding  the  tem- 
perature at  a  given  point;  for  example,  in  one  experiment  we  were 
able  to  cause  a  sample  of  pure  ammonium  benzoate  to  decompose 
at  185 0  by  holding  the  temperature  at  that  point  for  a  few  minutes 
(Ia  and  Alt  Table  III).  The  same  thing  has  been  shown  to  be 
true  of  other  compounds  which  decompose  on  heating. 
Further  investigation  of  binary  mixtures  containing  ammonium 
benzoate,  also  containing  other  compounds  which  decompose,  will 
probably  be  conducted  in  this  laboratory  in  connection  with  the 
investigation  of  melting  points  of  Pharmacopceial  compounds  pre- 
viously referred  to  in  this  article. 
Conclusions. — i.  The  results  of  the  simplified  distillation  method 
are  shown  to  be  sufficiently  accurate  for  the  quantitative  analysis 
of  Pharmacopceial  ammonium  benzoate  samples.  This  method  is 
better  adapted  to  the  quantitative  examination  of  ammonium  ben- 
zoate than  the  formaldehyde  method.  The  latter,  however,  is  no 
doubt  preferable  in  the  case  of  most  of  the  remaining  ammonium 
compounds  of  the  Pharmacopoeia. 
2.  The  melting  point  or  rather  decomposition  point  of  ammo- 
nium benzoate  is  shown  to  be  valueless  as  a  test  for  the  purity  of 
this  compound.  It  is  probable,  although  it  cannot  be  stated  with 
certainty,  that  with  all  other  Pharmacopceial  products  having  a 
melting  point  accompanied  by  decomposition  this  test  for  purity 
is,  as  in  the  case  of  ammonium  benzoate,  useless. 
3.  The  litmus  paper  test  is  shown  to  be  inadequate  for  the  detec- 
tion of  free  benzoic  acid  present  in  amounts  as  great  as  approxi- 
mately 10  per  cent.,  and  is  therefore  misleading. 
4.  The  Pharmacopceial  description  and  tests  for  ammonium  ben- 
zoate should  he  modified  by  the  elimination  of  the  litmus  paper  and 
melting-point  tests,  or  such  a  qualification  of  the  latter  that  will 
show  its  inapplicability  as  a  test  for  purity.  The  quantitative  esti- 
mation of  the  ammonia  by  distillation  may  be  briefly  described 
according  to  the  present  style  of  the  Pharmacopoeia  as  follows : 
The  ammonia  from  a  weighed  portion  of  about  0.5  Gm.  of  the 
sample,  dissolved  in  H20,  made  alkaline  with  50  c.c.  of  0.1 
N  NaOH,  is  distilled  into  50  c.c.  of  0.1  N  H2S04  and  the  excess  of 
the  latter  remaining  after  the  distillation  should  require  not  more 
than  14. 1  c.c.  of  0.1  N  NH4OH,  indicator  cochineal. 
