'ON  PREPARATION  OF  ALUMINIUM  FROM  CRYOLITE.  241 
There  is  also  great  loss  of  aluminium  caused  by  the  slow  cooling 
v>£  the  melted  mass  and  the  burning  of  the  pulverulent  metal, 
in  order  to  obviate  this,  another  mode  of  preparation  was 
adopted.  The  cryolite  (20  grin.)  was  ignited  strongly  in  a  gun- 
barrel,  through  which  was  passed  a  stream  of  hydrogen,  and 
then  the  vapor  of  8  grm.  of  sodium,  The  contents  of  the  tube 
were  cooled  in  an  atmosphere  of  hydrogen,  and  when  digested 
with  water,  to  dissolve  the  fluoride  of  sodium,  left  a  quantity  of 
a  black  powder  consisting  chiefly  of  iron. 
The  smal^  product  obtained  in  these  experiments  should  not 
deter  others  from  further  attempts  to  prepare  aluminium  from 
cryolite.  This  substance  can  now  be  obtained  at  such  a  low  price, 
and  sodium  is  likely  to  be  obtained  so  cheap,  as  to  render  desira- 
ble the  search  for  a  method  that  will  ensure  an  adequate  product. 
Moreover,  I  am  of  opinion  that  cryolite  is  better  adapted  for  the 
preparation  of  aluminium  than  any  compound  of  this  metal. 
Aluminium  has  not  yet  been  obtained  with  certainty  direct 
from  alumina.  Potassium  and  sodium  appear  to  effect  the  reduc- 
tion of  metallic  oxides  only  when  the  potash  or  soda  produced 
can  combine  with  the  unreduced  oxides.  Pure  potash  or  soda, 
the  characters  of  which  are  as  good  as  unknown,  do  not  appear 
to  be  formed.  Now,  as  alumina  combines  readily  with  the  alkalies 
forming  aluminates,  there  is  reason  to  believe  that  the  reduction 
might  be  effected  by  the  alkaline  metals. 
Even  if  this  should  prove  to  be  the  case,  cryolite  would  in  all 
probability  be  for  a  long  time  the  more  available  source  of  alumi- 
nium, unless  it  increased  very  much  in  price.  This  mineral  occurs 
in  a  state  of  singular  purity,  and  contains,  besides  aluminium, 
only  fluorine  and  sodium,  substances  which  do  not  interfere  with 
the  preparation  of  the  metal.  Alumina,  on  the  contrary,  is  rarely 
found  pure,  and  its  separation  on  a  large  scale,  from  substances 
that  would  be  prejudicial  to  the  production  of  the  metal,  would 
be  attended  with  great  difficulties. 
The  globules  of  aluminium  that  I  have  obtained  are  for  the 
most  part  so  malleable,  that  they  may  be  flattened  out  and  rolled 
into  foil  without  rending  at  the  edges.  They  have  a  strong 
metallic  lustre.  Some  pieces,  on  the  contrary,  which  were  found 
at  the  bottom  of  the  crucible,  sometimes  firmly  attached  to  it 
and  not  globular,  rent  at  the  sides  when  flattened.  They  also 
differ  somewhat  in  color  and  lustre.    They  are  evidently  not  so 
16 
