260  PRODUCING  HIGH  TEMPERATURES  BY  COAL  GAS  AND  AIR. 
the  mixture  of  air  and  gas  is  inflammable  is  barely  exceeded. 
In  spite  of  the  excess  of  air  the  combustion  is  then  very  incom- 
plete; the  hydrogen  burns,  but  the  carbon  only  forms  oxide  ;  the 
temperature  is  then  very  slightly  raised,  and  the  tube  bears  it 
without  injury.  Little  by  little  I  increase  the  air  and  the  pro- 
portion of  gas,  until  in  about  five  minutes  I  arrive  at  the  desired 
speed. 
For  heating  a  crucible  I  make  other  arrangements :  placing 
two  bricks  flatwise  together,  to  form  the  pedestal  of  the  furnace, 
in  the  centre  I  place  the  crucible  on  a  cheese ;  for  which  I  make 
a  verticle  covering  with  pieces  of  brick  of  equal  size  bound 
together  with  an  iron  wire.  This  covering  rests  on  four  blocks, 
so  arranged  as  to  have  between  it  and  the  pedestal  a  free  space 
of  three  to  four  millimetres.  This  verticle  envelope  I  cover  with 
a  brick  pierced  through  the  middle  to  receive  my  blow-pipe. 
Thus  I  heat  from  above,  the  flame  strikes  the  cover,  spreads 
over,  descends,  and  escapes  all  around  by  the  circular  outlet 
made  by  the  blocks.  * 
It  is  possible  to  vary  in  many  ways  the  form  of  the  flame  and 
that  of  the  coverings,  according  to  the  object  to  be  heated.  Sur- 
prising results  are  produced  by  this  mode  of  heating.  I  once 
melted,  in  twenty  minutes,  in  a  Paris  crucible,  a  piece  of  iron 
weighing  400  grammes,  and,  in  the  same  length  of  timeT  the  por- 
celain of  a  Bayeux  tube  was  transformed  into  transparent  glass. 
Nor  is  the  consumption  of  gas  excessive.  I  have  estimated  it 
approximately,  and  found  that,  to  heat  to  white  heat  during 
twenty  minutes,  a  porcelain  tube,  of  twenty  millimetres  by 
eighteen  centimetres  long,  about  250  litres  of  gas  were  con- 
sumed ;  while  from  400  to  500  were  required  to  melt  a  piece  of 
iron.  The  danger  of  melting  porcelain  tubes  necessitates  some 
precautions.  I  place  at  one  end  a  flask  with  the  bottom  black- 
ened, and  through  this  I  watch  the  effects  of  the  heat  on  the  por- 
celain ;  the  flask  is  tubulated  when  the  gas  is  required  to  circu- 
late in  the  tube.  On  the  first  appearance  of  any  distortion  in 
the  tube  I  diminish  the  current.  With  a  blow-pipe  and  a  furnace 
with  tubes,  the  pressure  of  air  corresponding  to  the  commence- 
ment of  fusion  of  porcelain  should  be  ascertained,  and  the  tem- 
perature kept  below  that  point.    This  pressure  limit  is  evidently 
