ON  ALUMINIUM. 
121 
sodium  as  a  curiosity  of  the  laboratory ;  it  was  then  produced  in 
small  quantities  by  heating  alumina  mixed  with  coal,  in  a  por- 
celain tube,  and  passing  over  it  a  current  of  dry  chlorine  gas, 
M.  Ste.  Claire  Deville  made  further  improvements  in  this  pro- 
cess, so  as  to  make  it  in  an  apparatus  as  large  as  a  gas  retort 
and  in  quantities  proportional,  at  a  price  of  twenty-five  cents  per 
pound.  To  produce  the  reaction  of  sodium  with  the  chloride  of 
aluminium  was  the  most  difficult  point  of  the  entire  process.  M. 
Ste.  Claire  Deville  used  for  the  reduction  the  distillation  of  the 
chloride  of  aluminium  over  the  sodium,  which  was  placed  in  trays 
of  copper  enclosed  in  a  tube.  The  temperature  developed  by 
the  reaction  is  very  great  if  the  current  of  the  chloride  of  alu- 
minium be  rapid ;  by  this  process  it  was  found  that  it  required 
at  least  ten  pounds  of  sodium  to  produce  one  pound  of  aluminium, 
(part  of  the  aluminium  produced  being  destroyed  at  its  forma- 
tion by  the  scoriae,)  when  by  the  theory  it  required  only  two  and 
a  half  pounds.  This  great  loss  of  sodium  and  the  difficulties  in 
conducting  this  reaction  on  a  large  scale,  were  very  great  objec- 
tions to  the  process. 
All  the  aluminium  at  the  Paris  Exhibition  was  made  by  this 
process,  and  it  was  from  a  portion  of  this  that  M.  Regnaulfc 
made  his  investigations,  and  in  which  he  found  copper  and  iron. 
The  copper  came  from  the  trays  in  which  the  reduction  was 
made.  The  presence  of  these  metals  in  small  quantities  will 
account  for  the  peculiar  physical  properties  which  he  ascribed  to 
aluminium. 
Circumstances  having  interrupted  M.  Ste.  Claire  Deville  in 
the  experiments  which  he  was  making  on  a  large  scale,  the  sub- 
ject rested  for  a  while  here.  In  the  meanwhile  Heinrich  Rose 
suggested  and  made  experiments  with  cryolite,  (a  fluoride  of 
aluminium  and  sodium,  (and  gave  his  views  that  this  mineral  was 
a  valuable  substance  from  which  to  produce  aluminium. 
Wohler  made  experiments  also  with  cryolite,  and  arrived  at 
conclusions  somewhat  similar  to  Heinrich  Rose.  They  both 
succeeded  in  producing  some  of  the  metal,  but  the  results  were 
not  entirely  satisfactory. 
M.  Ste.  Claire  Deville  again  resumed  his  experiments,  but  in- 
stead of  distilling  the  chloride  of  aluminium  on  the  metallic  sodium, 
as  in  his  first  experiments,  he  fused  in  a  crucible,  in  the  manner 
