PRODUCING  HIGH  TEMPERATURES  BY  COAL  GAS  AND  AIR.  259 
with  which  combustion  is  propagated  in  a  large  tube,  is  for  the 
theoretical  mixture  of  gas  and  air.  not  more  than  five  metres  the 
second.  If,  then,  the  speed  of  supply  in  mj  blowpipe  is  notably 
greater,  the  flame  would  not  be  able  to  back  so  as  to  burn  in  the 
interior  of  the  pipe.  Moreover,  an  explosion  under  such  condi- 
tions need  cause  no  alarm.  Nor  need  any  attention  be  paid  to 
the  fan  furnishing  the  air,  pressures  of  from  fifteen  to  twenty 
centimetres  of  water  being  quite  sufficient ;  but  great  care  must 
be  taken  to  provide  for  the  escape  of  the  products  of  combustion. 
I  use  M.  Enfer's  bellows,  which  I  regulate  by  conveying  the 
current  into  a  kind  of  gasometer  formed  by  a  large  receiver  fixed 
in  an  outer  cover  filled  with  water,  a  guage  indicating  the  pres- 
sure. The  gas  is  regulated  by  a  cock,  the  key  of  which  is  pro- 
vided with  a  long  handle,  so  that  very  slight  movements  may  be 
made.  The  mixture  is  at  its  highest  point  of  perfection  when 
two  very  slightly  different  positions  of  the  cock  give  by  turns 
oxidizing  and  deoxidizing  gases,  which  can  be  ascertained  by  ap- 
plying a  large  copper  wire  to  the  entrance  of  the  furnace. 
To  heat  a  porcelain  tube  to  white  heat  I  fix  at  the  end  of  the 
blowpipe  a  kind  of  flattened  funnel,  which  transforms  the  cyl- 
indrical jet  into  a  sheet  of  flame  ;  I  place  the  edge  of  the  funnel 
between  two  refractory  bricks  bound  together  by  iron  wires,  one 
of  them  having  been  filled  so  as  to  make  after  its  junction  with 
the  other  a  vacant  space,  forming  the  continuation  of  the  funnel, 
and  in  which  the  gaseous  sheet  spreads  out  until  it  escapes 
through  an  opening  from  11  tol8  centimetres  long  and  from  2  to 
3  millimetres  broad  ;  it  is  only  after  thus  escaping,  and  provided 
its  velocity  is  greater  than  that  above  mentioned,  that  it  burns. 
I  avoid  exposing  my  tube  too  near  the  escape,  as  the  porcelain 
would  be  melted  wherever  struck  directly  by  the  flame.  At  each 
side  and  at  the  two  ends  of  the  aperture  I  place  four  pieces  of 
brick,  confining  the  flame  in  a  space  from  one  to  two  centimetres 
broad  by  from  five  to  six  high.  A  little  above  I  place  my  tube, 
and  cover  it  with  some  conveniently-shaped  pieces  of  brick  ;  the 
lighted  gases  divided  by  the  tube  surround  it  and  reunite  above 
to  escape  by  a  longitudinal  aperture.  The  heating  should  of 
course  be  gradual  from  the  first.  I  commence  by  admitting  very 
little  air,  then  slowly  open  the  gas  cock  until  the  limit  at  which 
