196  Notes  on  Estimation  of  Nitroglycerine.  { Al\fyU  ri»ularm' 
determination  of  nitroglycerine  was  inaccurate,  and  this  would  be 
even  more  liable  to  error  in  anything  but  the  simplest  preparations. 
It  is,  therefore,  of  interest  to  note  the  comparisons  here  given  of 
nitroglycerine  content  as  determined  by  the  Scoville  method  and  by 
the  modified  Kjeldahl. 
Scoville 3  has  already  briefly  compared  these  values  and  the 
figures  here  given  lead  to  the  same  conclusions  as  his,  i.e.,  that  the 
results  obtained  by  these  methods  are  quite  alike,  and  that  .the  varia- 
tions are  due  to  the  difficulties  of  the  methods. 
For  the  analysis  of  tablets  containing  minute  quantities  there 
can  be  little  doubt  that  the  colorimetric  method  is  both  the  easier 
and  the  more  accurate.  But  in  the  case  of  the  stock  preparations 
such  as  the  concentrated  alcoholic  solutions  and  triturations,  the 
larger  quantities  of  nitroglycerine  under  analysis  make  it  possible 
to  obtain  closely  agreeing  results  by  the  modified  Kjeldahl  method. 
The  manufacturing  processes  can  be  closely  controlled  when  the 
stock  substances  are  standardized  by  the  Kjeldahl  and  the  finished 
product  by  the  colorimetric  method,  particularly  when  the  condi- 
tions are  so  standardized  that  a  definitely  known  loss  of  nitro- 
glycerine, previously  determined,  is  allowed  for.  Such  data  are 
given  in  the  table. 
It  will  be  observed,  that  under  the  conditions  employed  for  man- 
ufacture, the  losses  of  nitroglycerine  involved  have  been  reduced  to 
quite  narrow  limits,  averaging  about  five  per  cent  for  moulded 
tablets  and  about  eleven  per  cent  for  compressed  ones.  Our  find- 
ings on  the  compressed  tablets  "are  at  variance  with  the  losses  pre- 
viously recorded  4 ;  but  this  is  due  to  the  fact  that  the  tablets  are  pre- 
pared by  mixing  dry  granulations  instead  of  attempting  to  granulate 
the  mixture.  A  tablet  prepared  in  this  way  neither  discolors  nor 
loses  its  strength,  and  there  is  no  doubt  but  that  a  stable  standard 
product  results,  at  least  as  long  as  the  preparation  is  properly  pre- 
served. 
Experimental. 
For  the  colorimetric  analyses  a  method  equivalent  to  that  out- 
lined in  the  report 5  of  the  Official  Agricultural  Chemists  was  em- 
ployed. 
3  Lor.  cit. 
4  Amee.  Jour.  Pharm.,  79,  555,  1907. 
5  hoc.  cit. 
