ON  THE  MANUFACTURE  OF  ALUMINIUM. 
415 
the  vaporized  chloride  over  points  of  iron  heated  to  750°  Fahr., 
whereby  the  sesquichloride  of  iron  is  fixed  as  protochloride, 
whilst  the  aluminous  chloride  passes  out  and  condenses  in  color- 
less and  transparent  crystals. 
Sodium  may  now  be  prepared  with  facility  in  large  or  small 
vessels.  M.  Deville  has  carefully  studied  with  great  care  the  in- 
fluence of  temperature,  of  the  heating  surfaces,  and  the  quickness 
with  which  the  sodium  vapor  issues  from  the  apparatus,  and  he 
is  convinced  that  a  proper  regulation  of  the  heat,  and  the  diameter 
of  the  tubes  for  the  exit  of  the  sodium,  will  enable  him  to  produce 
this  metal  at  a  heat  as  low  as  that  of  melting  silver.  In  the 
experiments  already  made  the  heat  used  was  less  than  that  ne- 
cessary in  manufacturing  zinc.  The  author  is  now  seeking  a 
continuous  process  for  making  sodium,  as  he  finds  it  unnecessary 
to  redistil  the  metal,  it  being  pure  from  the  first  operation. 
In  relation  to  the  reaction  of  chloride  of  aluminium  with  the 
sodium,  it  is  effected  in  metallic  tubes,  which  in  shape  and  in 
process  of  management  have  not  yet  been  well  adapted  to  a 
manufacturing  scale,  yet  the  author  thinks  that  the  difficulties 
which  at  present  exist  will  soon  be  resolved. 
M.  Dumas  presented  to  the  Academy,  on  behalf  of  M.  Deville, 
some  large  and  beautiful  masses  of  chloride  of  aluminium,  of 
sodium,  and  of  metallic  aluminium  in  bars,  obtained  at  the  Javel 
works  during  the  course  of  experiments  there  made  at  the  ex- 
pense of  the  Emperor,  and  accompanied  them  with  the  following 
remarks  : 
The  process  of  making  chloride  of  aluminium  having  afforded 
it  to  the  extent  of  200  to  300  kilogrammes,  it  may  be  considered 
as  an  operation  susceptible  of  being  extended  on  a  manufacturing 
scale. 
M.  Deville's  process  for  sodium  yields  that  metal  with  great 
regularity  and  facility. 
The  materials  employed  in  the  manufacture  of  one  kilogramme 
of  aluminium — alumina  from  ammonia  alum,  chlorine,  charcoal, 
carbonate  of  soda  and  chalk — are  all  cheap,  being  worth  now  about 
32  francs,  whilst  the  price  of  sodium  before  was  1000  francs 
per  kilogramme,  at  which  rate  that  quantity  of  aluminium  would 
cost  3000  francs. 
M.  Deville's  experiments  not  only  demonstrated  the  possibility 
