Am.  Jour.  Pharm.  ) 
January,  1910.  J 
Ammonium  Benzoate, 
19 
the  two  curves  at  any  point  represents  the  range  over  which  the 
sample  was  melting  or  decomposing  (heating  at  the  rate  of  30  per 
minute). 
The  curves  show  clearly  that  the  variation  in  melting  point  or 
decomposing  point  between  a  sample  of  pure  ammonium  benzoate 
and  a  sample  containing  50  per  cent,  of  benzoic  acid  is  barely  sig- 
nificant. They  also  offer  a  striking  commentary  upon  the  general 
conception  of  the  melting  points  of  binary  mixtures  compared  to  the 
/ 
/  N 
f 
 T~* 
A 
Me 
tiny 
omj»os 
iiion  1 
/ 
A 
ft 
Ammo 
e 
f  Mitt 
'enzoa' 
uves  c 
e  and 
f 
Benzol 
cAcid 
— 
!  — 
J 
CiHfCMH  ^   <€        J^rcenta^e  Composition  o'f  Mixtures 
melting  point  of  either  constituent  in  pure  form.  The  almost  hori- 
zontal course  of  the  curves  from  100  per  cent,  to  50  per  cent, 
ammonium  benzoate  and  the  very  sharp  rise  at  94  per  cent,  to  96 
per  cent,  benzoic  acid  are  very  striking.  The  nature  of  the  curves 
would  doubtless  be  more  or  less  modified  by  further  experimenta- 
tion, and  perhaps  by  more  refined  methods,  but  for  the  principal 
purpose  of  this  work  w7e  believe  they  serve  just  as  well  as  they 
stand, — to  illustrate  the  unavailability  of  the  decomposition  point  as 
a  test  of  purity. 
A  further  indication  that  the  decomposition  point  is  an  unreliable 
test  for  purity  consists  in  the  fact  that  widely  varying  results  may 
