ON  LIQUOR  FEB, EI  PERACETATIS.  .  9 
cautions  must  be  observed,  which,  if  disregarded,  will  produce  a 
solution,  prone  to  change,  occasioning  the  precipitation  of  a 
large  portion  of  iron,  which  once  separated  cannot  be  redissolved 
again  by  the  further  addition  of  acetic  acid. 
In  the  first  place  it  is  indispensable  to  entirely  sesquioxidize  the 
iron  ;  if  a  portion  should  remain  in  the  state  of  protoxide,  a  pro- 
to-peroxide  would  afterwards  be  precipitated,  which  is  either  not 
entirely  soluble  in  cold  acetic  acid,  or  if  dissolved,  would  render 
the  solution  prone  to  change  through  its  affinity  for  oxygen. 
The  tersulphate  of  iron  as  well  as  the  ammonia  should  be 
considerably  diluted  with  water,  and  the  alkali  kept  in  excess. 
This  is  accomplished  by  using  more  than  an  equivalent  quantity 
of  ammonia,  and  by  pouring  the  diluted  iron  solution  into  the 
diluted  ammonia.  A  pure  hydrated  sesquioxide  is  then  precipi- 
tated free  from  basic  salt.  The  non-volatile  alkalies  are  not  so 
well  adapted  for  this  purpose,  because  the  precipitate  is  apt  to 
carry  down  a  portion  of  them  which  cannot  be  removed  by  sub- 
sequent washing.  The  precipitation,  moreover,  must  not  be  ac- 
complished too  rapidly,  in  order  to  avoid  the  partial  heating  of 
the  liquid  at  the  points  of  contact  with  the  precipitant,  whereby 
the  ferric  oxide  would  be  precipitated  in  a  different  state  of 
hydration,  being  then  insoluble  or  readily  separated  from  its  so- 
lution in  acetic  acid. 
The  hydrated  oxide  must  be  washed  with  water  and  expressed 
until  it  is  entirely  free  from  sulphate  of  ammonia ;  solution  of 
acetate  of  iron  is  gradually  precipitated  in  the  presence  of  other 
salts. 
The  acetic  acid  must  be  of  the  proper  strength.  If  stronger 
than  directed  a  neutral  peracetate  (Fe203,  3  ac)  will  be  formed; 
if  too  weak,  or  if  the  oxide  should  be  in  excess,  a  more  basic  salt 
is  obtained,  which  is  at  first  perfectly  soluble,  but  is  rapidly  de- 
composed by  precipitating  oxide  of  iron,  or  probably  a  very  basic 
acetate,  leaving  neutral  acetate  in  solution. 
Heat  must  not  be  applied,  either  during  the  precipitation  or 
washing  of  the  sesquioxide,  or  during  its  solution  in  acetic  acid. 
In  the  former  case  the  state  of  hydration  will  be  altered  as  stated 
before  ;  in  the  latter  case  the  oxide  will  at  first  dissolve  but  soon 
precipitate  wholly  or  in  part.    Solution  of  peracetate  of  iron 
