^"Aug"%f}T.'"-}     Development  of  the  Chemical  Arts,  363 
Berthier,  in  1822,  this  was  replaced  by  the  simple  ignition  of  manga- 
nese, and  finally  the  action  of  heat  upon  the  chlorate  of  potash.  The 
last-mentioned  process,  in  spite  of  its  costliness,  has  become  established 
for  laboratory  operations,  as  being  convenient  and  requiring  only  a 
small  supply  of  heat,  although  it  has  frequently  occasioned  explosions 
when  the  gas  was  being  too  rapidly  liberated.  To  prevent  such  acci- 
dents, it  has.  been  repeatedly  proposed  to  mix  manganese  with  the 
chlorate  of  potash.  Recent  accidents,  especially  a  fearful  explosion 
in  a  pharmaceutical  laboratory  in  Paris,  induced  Debray  and  Bourgoin^ 
to  make  known  the  precautions  used  in  Deville's  laboratory.  Man 
ganese,  or  the  red  manganoso-manganic  oxide  (MngO^),  which  is  more 
easily  obtained  in  a  state  of  purity,  is  mixed  with  the  chlorate  of  pot- 
ash in  equal  weights,  and  the  iron  retort  is  heated  in  a  furnace  filled 
with  fuel  in  such  a  manner  that  the  fire  may  be  kindled  at  the  top. 
Schwartzf  made  known  accidents  occasioned  by  the  use  of  manga- 
nese adulterated  with  lamp  black,  or  by  the  accidental  use  of  the  sul- 
phide of  antimony  instead  of  manganese,  and  he  therefore  very  justly 
recommends  that  oxygen  gas  mixture  should  be  previously  tested  by 
heating  a  portion  upon  platinum  foil.  Munckj  proposed  to  use  oxide 
of  iron  instead  of  manganese,  as  being  more  easily  distinguished. 
Scheele's  process — the  mutual  action  of  manganese  and  sulphuric 
acid — has  the  disadvantage  that  the  glass  is  often  broken  by  the  con- 
gelation of  the  manganous  sulphate.  To  prevent  this,  Wagner||  pro- 
poses to  use,  instead  of  sulphuric  acid,  bisulphate  of  soda.  An  easily 
fusible  double  salt  is  thus  formed  which  does  not  break  the  glass  as  it 
cools.  Pure  peroxide  of  manganese,  when  thus  treated,  evolves  18 
per  cent,  of  oxygen,  but  only  12  per  cent,  if  ignited,  when  it  is  con- 
verted into  sesquioxide.  Nevertheless,  the  latter  process  is  the  more 
economical.  Deville  and  Debray§  calculate  the  expense  according  to 
the  source  of  the  manganese,  as  follows  : 
Ten  kilos,  of  Manganese  Cost.  Price  of  i  cubic  metre  of  O. 
from  Francs.  F?-ancs. 
Romaneche        .        .        .  lo       .       .        .  .  4-86 
Spain      .       ,       ,       .  16  .       .       .  .  3-45 
Pyrenees     .       .       .        .  18        .       .       .  .  3*86 
Giessen           ...  27  ...  .  487 
Italy           ....  40       ...  .  5*98 
*  Debray  and  Boiirgoin,  "  Ber.  Chem,  Ges.  zu  Berlin,"  1870,  240. 
t  Schwartz,  "  Breslauer  Gewerbeblatt,"  1865,7. 
X  Munck,  "  Pohl's  Lehrbuch  der  Technologic  Wein,"  1865,  186. 
II  Wagner,  "  Jahresberichte,"  1866,  198. 
^  Deville  and  Debray,  "  Comptes  Rendus,"  li,  822. 
