PREPARATION  OF  METALLIC  COBALT. 
433 
loss,  3-3  per  cent.  260  parts  fractionally  distilled  with  steam 
gave — - 
First  product.  .  .  155  white  crystallized,  fusible  at  131°  F. 
Second  product.  .  32  yellowish  "  "  «  91°4F. 
Third  product  .  •     55  greenish  yellow,  consistence  of  honey. 
242 
Cocoa-nut  Oil  gave  equally  conclusive  results.  It  requires 
a  little  more  chloride  of  zinc,  in  consequence  of  the  large  quantity 
of  water  which  it  contains  or  produces. 
Oleic  Acid  300  parts  of  oleic  acid  from  a  candle  factory 
where  the  lime  saponification  is  practised,  treated  when  hot  with 
12  per  cent,  of  anhydrous  chloride  of  zinc,  furnished  on  distilla- 
tion a  solid  white  product  weighing  170,  and  fusible  at  89^.6  F., 
and  a  yellow  product  of  butyraceous  consistence  weighing  60. 
This  is  a  remarkable  fact,  and  clearly  shows  the  similarity  of  the 
action  of  chloride  of  zinc  and  sulphuric  acid.  By  means  of  sul- 
phuric acid,  25  to  30  per  cent,  of  solid  fatty  matter  is  obtained 
by  distillation  from  oleic  acid  produced  in  the  manufactories  of 
candles  where  lime-saponification  is  practised. — London  Ohem. 
Graz.  April  15,  1859,  from  Oomptes  Bendus, February  21, 1859- 
ON  THE  PREPARATION  OF  METALLIC  COBALT. 
By  William  Sharswood. 
The  attention  of  the  author  was  attracted  to  this  subject 
owing  to  a  necessity  of  obtaining  the  element  for  the  purpose  of 
making  researches  in  thermo-electricity. 
The  source  from  which  the  author  recommends  the  metal  to 
be  obtained  is  the  chloride  of  purpureocobalt,*  from  the  fact  of 
its  presenting  the  means  by  which  a  pure  chloride  of  cobalt 
can  be  obtained  with  comparatively  most  ease  and  certainty. 
The  formation  of  the  chloride  of  purpureocobalt  is  effected  by 
simply  oxidizing  the  chloride  of  cobalt  with  ammonia  by  ex- 
posure to  the  air. 
It  is  not  necessary  to  use  a  pure  chloride  of  cobalt  in  forming 
*  Researches  on  the  Ammonia-Cobalt  Bases,  bj  Walcott  Gibbs  and  Fred. 
Aug.  Genth,  Chemical  Gazette,  vol.  xv.  p.  188. 
