202 
Hydrogen  Peroxide. 
( Am.  Jour.  Pharru. 
\       June.  1S9S. 
plosions  reported  is  to  be  wondered  at.  Probably  this  may  be 
largely  accounted  for  by  the  fact  that  it  is  usually  the  custom  of  the 
manufacturer  to  replace  bottles  broken  spontaneously,  and  it  has 
necessarily  made  him  very  cautious  regarding  the  condition  of  the 
bottles  and  the  acidity  and  volume  strength  of  the  preparation  when 
it  leaves  his  hands. 
Quite  recently,  however,  the  writer's  attention  was  attracted  to  a 
new  property  of  this  article,  which  property,  it  is  believed,  has  never 
before  been  recorded. 
It  was  the  phenomenon  of  spontaneous  combustion,  caused  by 
hydrogen  peroxide  under  certain  conditions.  The  property  was  in- 
vestigated further,  and  experiments  verified  the  conclusion  that, 
under  certain  favorable  conditions,  hydrogen  peroxide  is  capable  oi 
causing  spontaneous  combustion. 
The  circumstances  attending  it  were  as  follows  :  About  April 
24th  the  writer  and  his  assistant  went  botanizing  in  New  Jersey. 
Neither  Rhus  toxicodendron  nor  Rhus  venenata  were  yet  in  leaf  nor 
flower,  although  many  of  the  plants  were  observed  and  avoided  on 
general  principles. 
The  next  day  the  assistant  discovered  that  his  right  wrist  was 
affected  by  Rhus  poisoning,  and  a  course  of  treatment  was  begun, 
using  peroxide  of  hydrogen,  which  seemed  to  act  very  favorably 
upon  the  eruption.  A  cheesecloth  bandage  was  wrapped  about  the 
wrist  and  kept  saturated  with  peroxide  by  pouring  on  a  small  quan- 
tity occasionally.  The  treatment  was  discontinued  in  the  evening, 
and  the  bandage  removed  before  retiring.  On  the  second  day  of  the 
treatment  a  rush  of  work  interfered  with  the  previous  practice  0: 
keeping  the  bandage  constantly  moistened,  and.  after  several  appli- 
cations during  the  earlier  part  of  the  morning,  the  matter  was  en- 
tirely forgotten.  Several  hours  later  an  odor  of  burning  clothes  and 
severe  pain  in  the  wrist  directed  the  patient's  attention  to  the  fact 
that  the  bandage  was  smouldering  and  was  alreadv  charred  black  in 
many  places.  Before  it  could  be  removed,  it  had  caused  several 
burns  on  the  wrist,  which  required  weeks  to  heal  and  which  will 
show  scars  for  several  years. 
The  experiment  was  afterward  repeated  by  saturating  the  same 
material  (cheesecloth)  with  peroxide  of  hydrogen  and  exposing  it  to 
the  heat  of  a  drying  oven,  meanwhile  observing  the  temperature  and 
action  of  the  peroxide.    Charring  took  place  in  all  cases,  but  in  no 
