68  The  Purity  Rubric  and  US.P.  Tests.  {^^SBT" 
0.5  N  HQ,  using  methyl  orange  as  indicator.  The  experience  in 
this  laboratory  has  shown  that  even  in  spite  of  the  greatest  care 
the  unburned  carbon  left  after  the  extraction  of  the  incinerated 
residue  retains  an  appreciable  amount  of  alkali,  and  therefore  in 
order  to  obtain  satisfactory  results  it  is  necessary  to  make  a  second 
ignition  of  this  unburned  and  extracted  carbon,  and  add  the  solution 
of  the  second  residue  to  that  of  the  first,  before  making  the  titration 
for  the  total  alkali.  The  modified  procedure  may  be  conveniently 
carried  out  as  follows : 
The  weighed  sample  is  ignited  thoroughly  in  a  platinum  dish, 
the  residue  extracted  with  hot  water,  and  the  solution  filtered 
through  an  ashless  filter,  the  unburned  carbon  washed  several  times, 
and  then  returned  together  with  the  filter  paper  to  the  platinum  dish 
and  ignited.  The  second  residue  is  dissolved  in  water  and  added 
to  the  filtered  extract  of  the  first  residue  and  the  solution  titrated 
with  0.5  N  HC1.  The  following  typical  results  indicate  the  necessity 
of  the  above  modification : 
C.c.  0.5  N  HCl  required.  Calculated  C6H5COOXa. 
Sample 
Wt.  used 
No. 
Gm. 
1st  Extract. 
2d  Extract. 
U.  S.  P. 
Modified. 
232 
1.0 
13-15 
0-35 
94-7 
974 
232 
I.O 
12.7 
O.85 
91-5 
97.6 
232 
I.O 
12.9 
0-5 
92.9 
96.5 
232 
I.O 
12.6 
0.8 
90.7 
96.5 
A  number  of  samples  of  sodium  benzoate  from  several  sources 
have  been  examined  by  the  above  modified  method,  but  none  con- 
tained the  99  per  cent,  pure  sodium  benzoate  required  by  the  purity 
rubric  of  the  Pharmacopoeia.  It  must  be  mentioned  that  the  experi- 
mental error  in  the  determination  could  be  reduced  considerably  by 
either  the  use  of  a  larger  sample  or  of  a  more  dilute  standard  acid. 
With  the  present  quantities,  an  error  of  0.1  c.c.  of  the  standard 
acid  corresponds  to  about  1.0  per  cent,  of  the  salt. 
As  indicated  above,  we  believe  that  the  foregoing  notes  serve  to 
illustrate  the  possibility  of  adapting  more  or  less  well-known  quan- 
titative methods  of  analysis  to  the  examination  of  Pharmacopceial 
compounds.  Such  simple  and  accurate  quantitative  methods 
together  with  qualitative  tests  for  impurities  of  a  serious  character 
will  give  a  ready  means  for  controlling  the  purity  rubric  of  the 
Pharmacopoeia,  and  raise  this  requirement  to  the  degree  of  impor- 
tance that  it  deserves. 
