IMPROVEMENTS  IN  THE  MANUFACTURE  OF  POTASSIUM.  71 
That  part  which  continues  in  a  state  of  vapor  when  it  reaches 
the  receiver,  and  which  afterwards  condenses,  is  also  lost ;  for 
the  researches  of  MM.  Mareska  and  Donny  have  shown  that  it 
passes  completely  to  the  state  of  a  black  substance  not  reducible 
by  heat. 
To  prevent  these  obstructions,  it  has  been  recommended  that 
the  intermediary  tube  should  be  made  as  short  as  possible,  and 
M.  Mitscherlich  adds  to  this  recommendation  the  useful  advice, 
to  maintain  it  at  a  red  heat  up  to  the  point  of  its  connection 
with  the  receiver,  in  the  hope  that  the  metal  would  condense  in 
the  less  heated  part  of  the  tube,  and  run,  drop  by  drop,  into  the 
receiver.  MM.  Mareska  and  Donny  observe,  that  by  limiting 
to  a  sufficient  extent  the  condensing  surface  of  the  pipe,  we  may 
effectively  diminish  the  space  in  which  these  obstructions  occur, 
and  the  difficulty  of  overcoming  them  by  means  of  the  iron  rod 
with  which  the  apparatus  is  furnished  ;  but  we  cannot,  by  these 
means,  entirely  remove  the  difficulties,  because,  however  limited 
may  be  the  portion  of  the  tube  where  the  metal  is  condensed,  we 
cannot  prevent  a  part  from  passing  to  a  solid  state,  remaining  so, 
and  by  becoming  carburetted,  forming  a  source  of  obstruction. 
Again,  nothing  is  more  difficult  than  to  maintain,  at  a  sufficient- 
ly low  temperature,  the  lower  part  of  a  metallic  tube,  the  remain- 
der of  which  is  kept  at  a  red  heat.  The  greater  part  of  the  po- 
tassium reduced,  escapes  in  the  state  of  vapor  the  whole  length 
of  the  tube,  and,  on  arriving  in  the  receptacle,  is  lost. 
From  these  facts,  MM.  Mareska  and  Donny  conclude,  that 
the  only  means  of  entirely  avoiding  the  obstruction,  and  diminish- 
ing the  loss  of  potassium,  consists  in  abandoning  the  idea  of  ef- 
fecting the  condensation  of  the  potassium  vapor  in  the  interme- 
diary tube  of  Brunner's  apparatus,  and  the  substitution  (for  this 
receiver  which  condenses  nothing,  or  which  condenses  only  car- 
buretted potassium)  of  a  receiver  in  which  the  vapor  solidifies  in 
the  metallic  state,  and  in  which  the  metal  can,  as  much  as  possi- 
ble, be  protected  from  the  alteration  which  carbonic  oxide  causes 
it  to  undergo.  The  condensation  no  longer  taking  place  in  the 
tube,  this  maybe  kept,  throughout  its  entire  length,  at  a  temper- 
ature at  which  the  action  of  carbonic  oxide  is  impossible. 
After  having  tried  a  great  number  of  receivers,  MM.  Mareska 
and  Donny  have  adopted  the  following  construction,  as  best  an- 
swering the  purpose  required  : — 
