Am.  Jour.  Pharm. ) 
August,  1911.  ; 
Estimation  of  Nitroglycerin. 
361 
of  Dr.  Rice  gave  this  process  an  immediate  standing,  and  it  has  since 
been  endorsed  by  several  writers,  notwithstanding  that  it  was 
pointed  out  as  early  as  1868  by  Tilberg  ^  a  Swedish  chemist  that  the 
reaction  between  potassium  hydroxide  and  nitroglycerin  is  not  a 
simple  but  a  very  complex  one,  and  results  in  nitrite,  cyanide, 
oxalate  and  formate  of  potassium  and  free  ammonia,  and  that  in 
1885  Hay  reiterated  this  fact.  The  endorsements  of  this  process, 
as  well  as  Dr.  Rice's  own  experiments  appear  to  be  based  on  the 
fact  that  expected  results  are  frequently  obtained  by  it,  but  without 
any  proof  that  expected  results  were  correct  results. 
In  19 10,  Berl  &  Delpy  ^  stated  that  when  cold  alcoholic  solution 
of  potassium  hydroxide  is  mixed  with  nitroglycerin,  and  the  tem- 
perature kept  under  25°  C.  for  some  time,  the  products  of  reaction 
are  potassiumi  nitrate,  nitrite,  cyanide,  oxalate,  mesoxalate  and 
formate,  aldehyde,  ammonia,  glyceryl  dinitrate  and  glyceryl  trini- 
trate—  (some  of  the  last  remaining  undecomposed) .  They  further 
state  that  6  molecules  of  potassium  hydroxide  are  required  to  de- 
compose one  of  nitroglycerin  and  that  the  reaction  is  not  complete 
in  the  cold. 
Here  then  there  is  one  chemist  (Hay)  who  says  that  one  molecule 
of  nitroglycerin  requires  5  molecules  of  alkali,  another  (Rice)  that 
3  molecules  of  alkali  are  required,  and  a  third  (Berl  &  Delpy)  that  6 
molecules  are  necessary. 
And  one  has  only  to  try  the  process,  varying  the  conditions  of 
heating,  the  time  of  standing,  and  the  temperature,  to  become  con- 
vinced that  the  results  are  of  no  value. 
Indeed  one  is  surprised  to  note  how  much  variation  in  results  is 
induced  by  slight  variations  in  the  process. 
In  1905,  Binz'  a  Swedish  chemist  proposed  to  estimate  nitro- 
glycerin by  saponifying  it  with  alcoholic  potash,  reducing  the  nitrate, 
cyanide,  etc.,  so  formed  to  ammonia  by  treating  the  liquid  with  nascent 
hydrogen  (formed  by  zinc  and  sulphuric  acid  added  to  the  liquid) 
then  estimating  the  ammonia,  after  distillation.  In  this  process  the 
nitrogen  is  first  converted  entirely  into  ammonia,  and  the  latter 
estimated.  This  appears  to  be  scientifically  sound,  and  if  conditious 
can  be  made  to  ensure  complete  reaction  without  loss  of  nitrogen 
in  any  form,  it  may  give  accurate  results.  The  writer  has  not  tried 
°  Proc.  A. Ph. A.,  1869,  page  242. 
Ber.  43  1421  thro.  Chem.  Al)st..  1010—2488. 
Year  Rook  Pliarni.,  1906,  page  53. 
