38 
FERRI  PYROPHOSPHAS. 
because  pyrophosphate  of  iron  is  soluble  in  solution  of  pyrophos- 
phate of  soda, — and  if  it  be  in  great  excess,  there  is  also  a  reac- 
tion involving  loss. 
With  the  proportions  by  weight  above  given,  there  is  no  loss, 
as  was  determined  by  a  careful  assay  of  the  magma.  The  re- 
sulting precipitate  of  pyro- phosphate  of  sesquioxide  of  iron, 
2Fe2  03,  3b  P05,  is  quite  insoluble  in  water,  and  may  therefore 
be  copiously  washed  without  loss.  It  is  in  a  very  finely  divided 
state,  and  so  light  that  it  settles  with  extreme  slowness,  so  that 
in  twenty-four  hours  it  still  occupies  one  half  of  the  washing 
bottle.  And  yet  the  washing  by  decantation,  as  directed,  is  the 
most  speedy  method,  because  the  precipitate  so  clogs  a  strainer 
as  to  render  washing  in  that  way  almost  impracticable.  The 
washing  directed  does  not  entirely  free  the  precipitate  from  sul- 
phate of  soda,  but  sufficiently  for  all  useful  purposes.  The 
magma  when  drained  as  directed,  having  been  turned  up  from 
the  surface  of  the  strainer  two  or  three  times  in  the  twenty-four 
hours,  weighs  thirty-four  parts.  Should  any  accident  occur, 
whereby  any  portion  of  the  magma  is  lost,  as  will  sometimes 
happen,  an  assay  of  the  remainder  should  be  made  by  drying 
and  heating  a  portion  of  it  carefully  weighed,  to  redness,  in  a 
platinum  capsule,  and  calculating  the  true  quantity  of  anhydrous 
salt,  by  applying  the  ratio  of  loss  obtained  in  the  assay  to  the 
whole  mass.  In  operating  on  the  large  scale,  when  pounds  are 
taken  for  parts,  the  writer  always  assays  the  magma  in  this  way, 
as  a  matter  of  precaution,  or  control.  The  results  obtained  in 
this  way  are  very  accurate. 
The  solution  of  citrate  of  ammonia  is  readily  made  without 
water,  in  the  way  directed,  the  point  of  saturation  being  attain- 
ed by  means  of  litmus  paper,  and  the  smell  of  ammonia  in  the 
solution.  The  ammonia  should  not  be  in  excess,  because  it  de- 
composes the  iron  salt  and  turns  the  solution  brown.  If,  by  ac- 
cident, much  excess  is  given  to  the  solution,  so  that  it  smells 
strongly  of  the  ammonia,  it  is  proper  to  drive  off  the  excess  by 
heating  before  the  solution  is  added  to  the  pyrophosphate. 
When  the  magma  and  solution  are  put  together  and  heated  with 
stirring,  a  bright  clear  olive  green  solution  is  speedily  obtained, 
provided  the  precipitation  has  been  effected  with  solutions  cooled 
down  below  60g.    When  precipitated  at  ordinary  summer  tem- 
