222 
ON  PYROXYLIN. 
not  be  reproduced  at  will.  Four  methods  of  decomposition  at 
100Q,  having  the  common  characteristic  of  the  disengagement 
of  nitrous  vapors  may  be  given  : — 
1.  The  pyroxylin  detonates  violently. 
2.  It  decomposes  without  detonating,  leaving  a  white,  pul- 
verulent, acid  residue,  partially  soluble  in  water,  containing  no 
nitrogen,  and  forming  about  half  the  weight  of  the  pyroxylin. 
3.  It  leaves  a  yellow,  amorphous,  inexplosible  residue,  par- 
tially soluble  in  water,  and  reducing,  like  glucose,  the  double 
tartrate  of  copper  and  potash. 
4.  It  gives  a  small  residue  (only  8  to  10  per  cent,  of  its 
weight),  and  a  black  matter,  in  appearance  like  charcoal.  In 
this  case  the  matrass  is  entirely  covered  with  a  yellow  powder, 
which  dissolves  in  alkali  with  considerable  disengagement  of 
ammonia  (this  matter  is  apparently  ulmate  of  ammonia).  From 
this  solution  acids  precipitate  a  dirty  yellow  body,  also  soluble 
in  alkalies.  The  charcoal-like  residue  disengages  ammonia  un- 
der the  action  of  potash.  This  production  of  ammonia  by  the 
simple  action  of  heat  from  a  matter  formed  of  nitric  acid  and 
cellulose  is  very  remarkable. 
Other  experiments  made  on  various  pyroxylin  at  90°  and 
then  at  80°  gave  exactly  the  same  results,  except  that  decom- 
position took  place  after  several  hours  instead  of  a  few  min- 
utes. 
At  60°,  and  even  at  55°,  pyroxylin  is  still  decomposed. 
After  a  few  days  the  matrass  becomes  full  of  dense  reddish  va- 
pors, and  the  same  non-nitrogenised  pulverulent  residue  of 
which  we  have  already  spoken  is  obtained.  No  combustion  was 
observed  in  these  latter  experiments. 
We  moreover  produced  detonation  by  putting  about  one 
gramme  of  pyroxylin  into  one  of  Gay-Lussac's  copper  stoves 
containing  oil  at  only  47°.  The  pyroxylin  thus  decomposed 
was  from  a  specimen  prepared  by  forty-eight  hours-'  immersion, 
and  washing  by  Lenk's  process. 
These  experiments  plainly  show  that,  contrary  to  General 
Lenk's  assertion,  his  pyroxylin  does  not  offer  more  resistance 
to  the  action  of  heat  than  that  of  Bouchet,  the  Austrian  silica- 
ted  pyroxylin  having  under  all  conditions  behaved  itself  in  the 
same  manner  as  the  others. 
